How to use tencent/Hy-MT2-30B-A3B with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # pip install "transformers<5.0.0" from transformers import pipeline pipe = pipeline("translation", model="tencent/Hy-MT2-30B-A3B")
# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("tencent/Hy-MT2-30B-A3B") model = AutoModelForCausalLM.from_pretrained("tencent/Hy-MT2-30B-A3B", device_map="auto")
请问论文中参评的外部模型使用的prompt是怎么样的？
该模型在翻译诗歌类的分行文字时，经常省略一半甚至更多不翻，特别是小语种比如越南语的诗时更是如此。更大的问题是即便要求符合原文的语用风格，出来的译文也是所谓的“创造性”翻译，外译中、中译外莫不如此。如果要与其讨论两句还会教训人，出来的文体风格就是目前社交媒体上那种风格的文字，外译中时（特别是文学类素材）极大几率出来的是所谓“美文”。
· Sign up or log in to comment
