| 1 | # SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved. |
| 2 | # SPDX-License-Identifier: Apache-2.0 |
| 3 | # |
| 4 | # Licensed under the Apache License, Version 2.0 (the "License"); |
| 5 | # you may not use this file except in compliance with the License. |
| 6 | # You may obtain a copy of the License at |
| 7 | # |
| 8 | # http://www.apache.org/licenses/LICENSE-2.0 |
| 9 | # |
| 10 | # Unless required by applicable law or agreed to in writing, software |
| 11 | # distributed under the License is distributed on an "AS IS" BASIS, |
| 12 | # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 13 | # See the License for the specific language governing permissions and |
| 14 | # limitations under the License. |
| 15 | from .baichuan.model import BaichuanForCausalLM |
| 16 | from .bert.model import ( |
| 17 | BertForQuestionAnswering, |
| 18 | BertForSequenceClassification, |
| 19 | BertModel, |
| 20 | RobertaForQuestionAnswering, |
| 21 | RobertaForSequenceClassification, |
| 22 | RobertaModel, |
| 23 | ) |
| 24 | from .bloom.model import BloomForCausalLM, BloomModel |
| 25 | from .chatglm.config import ChatGLMConfig |
| 26 | from .chatglm.model import ChatGLMForCausalLM, ChatGLMModel |
| 27 | from .cogvlm.config import CogVLMConfig |
| 28 | from .cogvlm.model import CogVLMForCausalLM |
| 29 | from .commandr.model import CohereForCausalLM |
| 30 | from .dbrx.config import DbrxConfig |
| 31 | from .dbrx.model import DbrxForCausalLM |
| 32 | from .deepseek_v1.model import DeepseekForCausalLM |
| 33 | from .deepseek_v2.model import DeepseekV2ForCausalLM |
| 34 | from .dit.model import DiT |
| 35 | from .eagle.model import EagleForCausalLM |
| 36 | from .enc_dec.model import DecoderModel, EncoderModel, WhisperEncoder |
| 37 | from .f5tts.model import F5TTS |
| 38 | from .falcon.config import FalconConfig |
| 39 | from .falcon.model import FalconForCausalLM, FalconModel |
| 40 | from .gemma.config import GEMMA2_ARCHITECTURE, GEMMA_ARCHITECTURE, GemmaConfig |
| 41 | from .gemma.model import GemmaForCausalLM |
| 42 | from .gpt.config import GPTConfig |
| 43 | from .gpt.model import GPTForCausalLM, GPTModel |
| 44 | from .gptj.config import GPTJConfig |
| 45 | from .gptj.model import GPTJForCausalLM, GPTJModel |
| 46 | from .gptneox.model import GPTNeoXForCausalLM, GPTNeoXModel |
| 47 | from .grok.model import GrokForCausalLM |
| 48 | from .llama.config import LLaMAConfig |
| 49 | from .llama.model import LLaMAForCausalLM, LLaMAModel |
| 50 | from .mamba.model import MambaForCausalLM |
| 51 | from .medusa.config import MedusaConfig |
| 52 | from .medusa.model import MedusaForCausalLm |
| 53 | from .mllama.model import MLLaMAModel |
| 54 | from .modeling_utils import PretrainedConfig, PretrainedModel, SpeculativeDecodingMode |
| 55 | from .mpt.model import MPTForCausalLM, MPTModel |
| 56 | from .nemotron_nas.model import DeciLMForCausalLM |
| 57 | from .opt.model import OPTForCausalLM, OPTModel |
| 58 | from .phi.model import PhiForCausalLM, PhiModel |
| 59 | from .phi3.model import Phi3ForCausalLM, Phi3Model |
| 60 | from .qwen.model import QWenForCausalLM |
| 61 | from .recurrentgemma.model import RecurrentGemmaForCausalLM |
| 62 | from .redrafter.model import ReDrafterForCausalLM |
| 63 | |
| 64 | |
| 65 | __all__ = [ |
| 66 | "BertModel", |
| 67 | "BertForQuestionAnswering", |
| 68 | "BertForSequenceClassification", |
| 69 | "RobertaModel", |
| 70 | "RobertaForQuestionAnswering", |
| 71 | "RobertaForSequenceClassification", |
| 72 | "BloomModel", |
| 73 | "BloomForCausalLM", |
| 74 | "DiT", |
| 75 | "DeepseekForCausalLM", |
| 76 | "FalconConfig", |
| 77 | "DeepseekV2ForCausalLM", |
| 78 | "FalconForCausalLM", |
| 79 | "FalconModel", |
| 80 | "GPTConfig", |
| 81 | "GPTModel", |
| 82 | "GPTForCausalLM", |
| 83 | "OPTForCausalLM", |
| 84 | "OPTModel", |
| 85 | "LLaMAConfig", |
| 86 | "LLaMAForCausalLM", |
| 87 | "LLaMAModel", |
| 88 | "MedusaConfig", |
| 89 | "MedusaForCausalLm", |
| 90 | "ReDrafterForCausalLM", |
| 91 | "GPTJConfig", |
| 92 | "GPTJModel", |
| 93 | "GPTJForCausalLM", |
| 94 | "GPTNeoXModel", |
| 95 | "GPTNeoXForCausalLM", |
| 96 | "PhiModel", |
| 97 | "PhiConfig", |
| 98 | "Phi3Model", |
| 99 | "Phi3Config", |
| 100 | "PhiForCausalLM", |
| 101 | "Phi3ForCausalLM", |
| 102 | "ChatGLMConfig", |
| 103 | "ChatGLMForCausalLM", |
| 104 | "ChatGLMModel", |
| 105 | "BaichuanForCausalLM", |
| 106 | "QWenConfigQWenForCausalLM", |
| 107 | "QWenModel", |
| 108 | "EncoderModel", |
| 109 | "DecoderModel", |
| 110 | "PretrainedConfig", |
| 111 | "PretrainedModel", |
| 112 | "WhisperEncoder", |
| 113 | "MambaForCausalLM", |
| 114 | "MambaConfig", |
| 115 | "MPTForCausalLM", |
| 116 | "MPTModel", |
| 117 | "SkyworkForCausalLM", |
| 118 | "GemmaConfig", |
| 119 | "GemmaForCausalLM", |
| 120 | "DbrxConfig", |
| 121 | "DbrxForCausalLM", |
| 122 | "RecurrentGemmaForCausalLM", |
| 123 | "CogVLMConfig", |
| 124 | "CogVLMForCausalLM", |
| 125 | "EagleForCausalLM", |
| 126 | "SpeculativeDecodingMode", |
| 127 | "CohereForCausalLM", |
| 128 | "MLLaMAModel", |
| 129 | "F5TTS", |
| 130 | ] |
| 131 | |
| 132 | MODEL_MAP = { |
| 133 | "GPT2LMHeadModel": GPTForCausalLM, |
| 134 | "GPT2LMHeadCustomModel": GPTForCausalLM, |
| 135 | "GPTBigCodeForCausalLM": GPTForCausalLM, |
| 136 | "Starcoder2ForCausalLM": GPTForCausalLM, |
| 137 | "FuyuForCausalLM": GPTForCausalLM, |
| 138 | "Kosmos2ForConditionalGeneration": GPTForCausalLM, |
| 139 | "JAISLMHeadModel": GPTForCausalLM, |
| 140 | "GPTForCausalLM": GPTForCausalLM, |
| 141 | "NemotronForCausalLM": GPTForCausalLM, |
| 142 | "OPTForCausalLM": OPTForCausalLM, |
| 143 | "BloomForCausalLM": BloomForCausalLM, |
| 144 | "RWForCausalLM": FalconForCausalLM, |
| 145 | "FalconForCausalLM": FalconForCausalLM, |
| 146 | "PhiForCausalLM": PhiForCausalLM, |
| 147 | "Phi3ForCausalLM": Phi3ForCausalLM, |
| 148 | "Phi3VForCausalLM": Phi3ForCausalLM, |
| 149 | "Phi3SmallForCausalLM": Phi3ForCausalLM, |
| 150 | "PhiMoEForCausalLM": Phi3ForCausalLM, |
| 151 | "MambaForCausalLM": MambaForCausalLM, |
| 152 | "GPTNeoXForCausalLM": GPTNeoXForCausalLM, |
| 153 | "GPTJForCausalLM": GPTJForCausalLM, |
| 154 | "MPTForCausalLM": MPTForCausalLM, |
| 155 | "GLMModel": ChatGLMForCausalLM, |
| 156 | "ChatGLMModel": ChatGLMForCausalLM, |
| 157 | "ChatGLMForCausalLM": ChatGLMForCausalLM, |
| 158 | "LlamaForCausalLM": LLaMAForCausalLM, |
| 159 | "ExaoneForCausalLM": LLaMAForCausalLM, |
| 160 | "MistralForCausalLM": LLaMAForCausalLM, |
| 161 | "MixtralForCausalLM": LLaMAForCausalLM, |
| 162 | "ArcticForCausalLM": LLaMAForCausalLM, |
| 163 | "Grok1ModelForCausalLM": GrokForCausalLM, |
| 164 | "InternLMForCausalLM": LLaMAForCausalLM, |
| 165 | "InternLM2ForCausalLM": LLaMAForCausalLM, |
| 166 | "MedusaForCausalLM": MedusaForCausalLm, |
| 167 | "ReDrafterForCausalLM": ReDrafterForCausalLM, |
| 168 | "BaichuanForCausalLM": BaichuanForCausalLM, |
| 169 | "BaiChuanForCausalLM": BaichuanForCausalLM, |
| 170 | "SkyworkForCausalLM": LLaMAForCausalLM, |
| 171 | GEMMA_ARCHITECTURE: GemmaForCausalLM, |
| 172 | GEMMA2_ARCHITECTURE: GemmaForCausalLM, |
| 173 | "QWenLMHeadModel": QWenForCausalLM, |
| 174 | "QWenForCausalLM": QWenForCausalLM, |
| 175 | "Qwen2ForCausalLM": QWenForCausalLM, |
| 176 | "Qwen2MoeForCausalLM": QWenForCausalLM, |
| 177 | "Qwen2ForSequenceClassification": QWenForCausalLM, |
| 178 | "Qwen2VLForConditionalGeneration": QWenForCausalLM, |
| 179 | "WhisperEncoder": WhisperEncoder, |
| 180 | "EncoderModel": EncoderModel, |
| 181 | "DecoderModel": DecoderModel, |
| 182 | "DbrxForCausalLM": DbrxForCausalLM, |
| 183 | "RecurrentGemmaForCausalLM": RecurrentGemmaForCausalLM, |
| 184 | "CogVLMForCausalLM": CogVLMForCausalLM, |
| 185 | "DiT": DiT, |
| 186 | "DeepseekForCausalLM": DeepseekForCausalLM, |
| 187 | "DeciLMForCausalLM": DeciLMForCausalLM, |
| 188 | "DeepseekV2ForCausalLM": DeepseekV2ForCausalLM, |
| 189 | "EagleForCausalLM": EagleForCausalLM, |
| 190 | "CohereForCausalLM": CohereForCausalLM, |
| 191 | "MllamaForConditionalGeneration": MLLaMAModel, |
| 192 | "BertForQuestionAnswering": BertForQuestionAnswering, |
| 193 | "BertForSequenceClassification": BertForSequenceClassification, |
| 194 | "BertModel": BertModel, |
| 195 | "RobertaModel": RobertaModel, |
| 196 | "RobertaForQuestionAnswering": RobertaForQuestionAnswering, |
| 197 | "RobertaForSequenceClassification": RobertaForSequenceClassification, |
| 198 | "F5TTS": F5TTS, |
| 199 | } |
| 200 |