Add m2m_100 model (#1068)

Author: liuchuting

mindone/transformers/__init__.py

@@ -260,6 +260,7 @@
 )
 from .models.llama import LlamaForCausalLM, LlamaForSequenceClassification, LlamaModel, LlamaPreTrainedModel
 from .models.llava import LlavaConfig, LlavaForConditionalGeneration
+from .models.m2m_100 import M2M100ForConditionalGeneration, M2M100Model, M2M100PreTrainedModel
 from .models.megatron_bert import (
     MegatronBertForCausalLM,
     MegatronBertForMaskedLM,

mindone/transformers/models/__init__.py

@@ -49,6 +49,7 @@
     levit,
     llama,
     llava,
+    m2m_100,
     megatron_bert,
     minicpm4,
     mistral,

mindone/transformers/models/auto/configuration_auto.py

@@ -69,6 +69,7 @@
         ("imagegpt", "ImageGPTConfig"),
         ("led", "LEDConfig"),
         ("levit", "LevitConfig"),
+        ("m2m_100", "M2M100Config"),
         ("llama", "LlamaConfig"),
         ("persimmon", "PersimmonConfig"),
         ("fuyu", "FuyuConfig"),
@@ -111,6 +112,7 @@
         ("albert", "ALBERT"),
         ("aria", "Aria"),
         ("aria_text", "AriaText"),
+        ("m2m_100", "M2M100"),
         ("bert", "BERT"),
         ("bart", "BART"),
         ("camembert", "CamemBERT"),

mindone/transformers/models/auto/modeling_auto.py

@@ -51,6 +51,7 @@
         ("deberta-v2", "DebertaV2Model"),
         ("dpt", "DPTModel"),
         ("gemma", "GemmaModel"),
+        ("m2m_100", "M2M100Model"),
         ("gemma2", "Gemma2Model"),
         ("persimmon", "PersimmonModel"),
         ("glm", "GlmModel"),
@@ -135,6 +136,7 @@
         # Model with LM heads mapping
         ("albert", "AlbertForMaskedLM"),
         ("bart", "BartForConditionalGeneration"),
+        ("m2m_100", "M2M100ForConditionalGeneration"),
         ("bert", "BertForMaskedLM"),
         ("deberta", "DebertaForMaskedLM"),
         ("deberta-v2", "DebertaV2ForMaskedLM"),
@@ -328,6 +330,7 @@
         # Model for Seq2Seq Causal LM mapping
         ("bart", "BartForConditionalGeneration"),
         ("led", "LEDForConditionalGeneration"),
+        ("m2m_100", "M2M100ForConditionalGeneration"),
         ("mt5", "MT5ForConditionalGeneration"),
         ("qwen2_audio", "Qwen2AudioForConditionalGeneration"),
         ("t5", "T5ForConditionalGeneration"),

mindone/transformers/models/m2m_100/__init__.py

@@ -0,0 +1,17 @@
+# 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_m2m_100 import M2M100ForConditionalGeneration, M2M100Model, M2M100PreTrainedModel