added cbs to notebooks, made copy-paste error fix in generation_utils (#16246)

notebooks/README.md

@@ -57,6 +57,7 @@ You can open any page of the documentation as a notebook in colab (there is a bu
 | [How to fine-tune a model on audio classification](https://github.com/huggingface/notebooks/blob/master/examples/audio_classification.ipynb)| Show how to preprocess the data and fine-tune a pretrained Speech model on Keyword Spotting | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/audio_classification.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/master/examples/audio_classification.ipynb)| 
 | [How to train a language model from scratch](https://github.com/huggingface/blog/blob/master/notebooks/01_how_to_train.ipynb)| Highlight all the steps to effectively train Transformer model on custom data | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/blog/blob/master/notebooks/01_how_to_train.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/huggingface/blog/blob/master/notebooks/01_how_to_train.ipynb)| 
 | [How to generate text](https://github.com/huggingface/blog/blob/master/notebooks/02_how_to_generate.ipynb)| How to use different decoding methods for language generation with transformers | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/blog/blob/master/notebooks/02_how_to_generate.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/huggingface/blog/blob/master/notebooks/02_how_to_generate.ipynb)| 
+| [How to generate text (with constraints)](https://github.com/huggingface/blog/blob/master/notebooks/53_constrained_beam_search.ipynb)| How to guide language generation with user-provided constraints | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/blog/blob/master/notebooks/53_constrained_beam_search.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/huggingface/blog/blob/master/notebooks/53_constrained_beam_search.ipynb)| 
 | [How to export model to ONNX](https://github.com/huggingface/notebooks/blob/master/examples/onnx-export.ipynb)| Highlight how to export and run inference workloads through ONNX |
 | [How to use Benchmarks](https://github.com/huggingface/notebooks/blob/master/examples/benchmark.ipynb)| How to benchmark models with transformers | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/benchmark.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/huggingface/notebooks/blob/master/examples/benchmark.ipynb)| 
 | [Reformer](https://github.com/huggingface/blog/blob/master/notebooks/03_reformer.ipynb)| How Reformer pushes the limits of language modeling | [![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/patrickvonplaten/blog/blob/master/notebooks/03_reformer.ipynb)| [![Open in AWS Studio](https://studiolab.sagemaker.aws/studiolab.svg)](https://studiolab.sagemaker.aws/import/github/patrickvonplaten/blog/blob/master/notebooks/03_reformer.ipynb)| 

src/transformers/generation_utils.py

@@ -3075,7 +3075,7 @@ class GenerationMixin:
         ...     )
         ... }
 
-        >>> constraint_str = "sind"
+        >>> constraint_str = "Sie"
         >>> constraint_token_ids = tokenizer.encode(constraint_str)[:-1]  # slice to remove eos token
         >>> constraints = [PhrasalConstraint(token_ids=constraint_token_ids)]
 
@@ -3175,10 +3175,6 @@ class GenerationMixin:
                 continue  # don't waste resources running the code we don't need
 
             next_token_logits = outputs.logits[:, -1, :]
-
-            # hack: adjust tokens for Marian. For Marian we have to make sure that the `pad_token_id`
-            # cannot be generated both before and after the `nn.functional.log_softmax` operation.
-            next_token_logits = outputs.logits[:, -1, :]
             # hack: adjust tokens for Marian. For Marian we have to make sure that the `pad_token_id`
             # cannot be generated both before and after the `nn.functional.log_softmax` operation.
             next_token_logits = self.adjust_logits_during_generation(next_token_logits, cur_len=cur_len)