Merge pull request #310 from Xiaoxue-xx/2.0.0

fix and add instructions

Author: StevenTang1998

instructions/Multilingual Models/M2M_100.md

@@ -1,6 +1,6 @@
 ## M2M_100
 
-You can fine-tune a NLLB model from HuggingFace through ``model=NLLB``, ``model_path=<hf-or-local-path>``, ``dataset=<dataset-name>``. Specifically, M2M is a Multilingual Model. You can set source language and target language with ``tokenizer_kwargs={'src_lang': 'source language', 'tgt_lang': 'target_language'}``.
+You can fine-tune a M2M_100 model from HuggingFace through ``model=M2M_100``, ``model_path=<hf-or-local-path>``, ``dataset=<dataset-name>``. Specifically, M2M is a Multilingual Model. You can set source language and target language with ``tokenizer_kwargs={'src_lang': 'source language', 'tgt_lang': 'target_language'}``.
 
 Example usage:
 

instructions/Seq2Seq/MASS.md

@@ -0,0 +1,12 @@
+## MASS
+
+You can fine-tune a MASS model from HuggingFace through ``model=MASS``, ``model_path=<hf-or-local-path>``, ``dataset=<dataset-name>``. 
+
+Example usage:
+
+```bash
+python run_textbox.py \
+    --model=MASS \
+    --model_path=RUCAIBox/mass-base-uncased \
+    --dataset=samsum
+```

instructions/Seq2Seq/UniLM.md

@@ -1,6 +1,6 @@
 ## UniLM
 
-You can fine-tune a T5 model from HuggingFace through ``model=T5``, ``model_path=<hf-or-local-path>``, ``dataset=<dataset-name>``. 
+You can fine-tune a UniLM model from HuggingFace through ``model=UniLM``, ``model_path=<hf-or-local-path>``, ``dataset=<dataset-name>``. 
 
 Example usage: