Add Mvp model (#1085)

* Add mvp model

* Add mvp model

Author: liuchuting

mindone/transformers/__init__.py

@@ -317,6 +317,14 @@
     MT5Model,
     MT5PreTrainedModel,
 )
+from .models.mvp import (
+    MvpForCausalLM,
+    MvpForConditionalGeneration,
+    MvpForQuestionAnswering,
+    MvpForSequenceClassification,
+    MvpModel,
+    MvpPreTrainedModel,
+)
 from .models.paligemma import PaliGemmaForConditionalGeneration, PaliGemmaPreTrainedModel
 from .models.persimmon import (
     PersimmonForCausalLM,

mindone/transformers/models/__init__.py

@@ -56,6 +56,7 @@
     mixtral,
     mobilebert,
     mpt,
+    mvp,
     paligemma,
     persimmon,
     phi,

mindone/transformers/models/auto/configuration_auto.py

@@ -98,6 +98,7 @@
         ("t5", "T5Config"),
         ("umt5", "UMT5Config"),
         ("wav2vec2", "Wav2Vec2Config"),
+        ("mvp", "MvpConfig"),
         ("whisper", "WhisperConfig"),
         ("xlm-roberta", "XLMRobertaConfig"),
         ("xlm-roberta-xl", "XLMRobertaXLConfig"),
@@ -118,6 +119,7 @@
         ("camembert", "CamemBERT"),
         ("bit", "BiT"),
         ("blip", "BLIP"),
+        ("mvp", "MVP"),
         ("blip-2", "BLIP-2"),
         ("chameleon", "Chameleon"),
         ("clap", "CLAP"),

mindone/transformers/models/auto/modeling_auto.py

@@ -38,6 +38,7 @@
         ("bert", "BertModel"),
         ("bart", "BartModel"),
         ("camembert", "CamembertModel"),
+        ("mvp", "MvpModel"),
         ("bit", "BitModel"),
         ("blip", "BlipModel"),
         ("blip-2", "Blip2Model"),
@@ -110,6 +111,7 @@
         ("albert", "AlbertForPreTraining"),
         ("bart", "BartForConditionalGeneration"),
         ("camembert", "CamembertForMaskedLM"),
+        ("mvp", "MvpForConditionalGeneration"),
         ("bert", "BertForPreTraining"),
         ("gpt2", "GPT2LMHeadModel"),
         ("gemma3", "Gemma3ForConditionalGeneration"),
@@ -135,6 +137,7 @@
     [
         # Model with LM heads mapping
         ("albert", "AlbertForMaskedLM"),
+        ("mvp", "MvpForConditionalGeneration"),
         ("bart", "BartForConditionalGeneration"),
         ("m2m_100", "M2M100ForConditionalGeneration"),
         ("bert", "BertForMaskedLM"),
@@ -162,6 +165,7 @@
         ("aria_text", "AriaTextForCausalLM"),
         ("bart", "BartForCausalLM"),
         ("camembert", "CamembertForCausalLM"),
+        ("mvp", "MvpForCausalLM"),
         ("bert", "BertLMHeadModel"),
         ("bert-generation", "BertGenerationDecoder"),
         ("gemma", "GemmaForCausalLM"),
@@ -278,6 +282,7 @@
 MODEL_FOR_MASKED_LM_MAPPING_NAMES = OrderedDict(
     [
         # Model for Masked LM mapping
+        ("mvp", "MvpForConditionalGeneration"),
         ("albert", "AlbertForMaskedLM"),
         ("bart", "BartForConditionalGeneration"),
         ("bert", "BertForMaskedLM"),
@@ -331,6 +336,7 @@
         ("bart", "BartForConditionalGeneration"),
         ("led", "LEDForConditionalGeneration"),
         ("m2m_100", "M2M100ForConditionalGeneration"),
+        ("mvp", "MvpForConditionalGeneration"),
         ("mt5", "MT5ForConditionalGeneration"),
         ("qwen2_audio", "Qwen2AudioForConditionalGeneration"),
         ("t5", "T5ForConditionalGeneration"),
@@ -352,6 +358,7 @@
         ("bart", "BartForSequenceClassification"),
         ("camembert", "CamembertForSequenceClassification"),
         ("bert", "BertForSequenceClassification"),
+        ("mvp", "MvpForSequenceClassification"),
         ("roberta", "RobertaForSequenceClassification"),
         ("deberta", "DebertaForSequenceClassification"),
         ("deberta-v2", "DebertaV2ForSequenceClassification"),
@@ -386,6 +393,7 @@
         ("bart", "BartForQuestionAnswering"),
         ("bert", "BertForQuestionAnswering"),
         ("camembert", "CamembertForQuestionAnswering"),
+        ("mvp", "MvpForQuestionAnswering"),
         ("roberta", "RobertaForQuestionAnswering"),
         ("deberta", "DebertaForQuestionAnswering"),
         ("deberta-v2", "DebertaV2ForQuestionAnswering"),

mindone/transformers/models/mvp/__init__.py

@@ -0,0 +1,24 @@
+# Copyright 2024 The HuggingFace Team. All rights reserved.
+#
+# This code is adapted from https://github.com/huggingface/transformers
+# with modifications to run transformers on mindspore.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from .modeling_mvp import (
+    MvpForCausalLM,
+    MvpForConditionalGeneration,
+    MvpForQuestionAnswering,
+    MvpForSequenceClassification,
+    MvpModel,
+    MvpPreTrainedModel,
+)