Nicolas Patry c2d0ffec8c Adding a new return_full_text parameter to TextGenerationPipeline. (#9852) ...

* Adding a new `return_full_text` parameter to TextGenerationPipeline.

For text-generation, it's sometimes used as prompting text.
In that context, prefixing `generated_text` with the actual input
forces the caller to take an extra step to remove it.

The proposed change adds a new parameter (for backward compatibility).
`return_full_text` that enables the caller to prevent adding the prefix.

* Doc quality.

2021-01-29 10:27:32 +01:00

..

benchmark

Transformers fast import part 2 (#9446)

2021-01-07 09:36:14 -05:00

commands

Add t5 convert to transformers-cli (#9654)

2021-01-20 09:34:27 -05:00

data

New run_seq2seq script (#9605)

2021-01-19 15:22:17 -05:00

models

Fix computation of attention_probs when head_mask is provided. (#9853)

2021-01-28 06:11:52 -05:00

pipelines

Adding a new return_full_text parameter to TextGenerationPipeline. (#9852)

2021-01-29 10:27:32 +01:00

utils

ConvBERT Model (#9717)

2021-01-27 03:20:09 -05:00

__init__.py

ConvBERT Model (#9717)

2021-01-27 03:20:09 -05:00

activations_tf.py

Fix mixed precision in TF models (#9163)

2021-01-21 07:00:11 -05:00

activations.py

Copyright (#8970)

2020-12-07 18:36:34 -05:00

configuration_utils.py

Add flags to return scores, hidden states and / or attention weights in GenerationMixin (#9150)

2021-01-06 17:11:42 +01:00

convert_graph_to_onnx.py

Transformers fast import part 2 (#9446)

2021-01-07 09:36:14 -05:00

convert_pytorch_checkpoint_to_tf2.py

Transformers fast import part 2 (#9446)

2021-01-07 09:36:14 -05:00

convert_slow_tokenizer.py

ConvBERT Model (#9717)

2021-01-27 03:20:09 -05:00

convert_slow_tokenizers_checkpoints_to_fast.py

Transformers fast import part 2 (#9446)

2021-01-07 09:36:14 -05:00

convert_tf_hub_seq_to_seq_bert_to_pytorch.py

Transformers fast import part 2 (#9446)

2021-01-07 09:36:14 -05:00

dependency_versions_check.py

Fast transformers import part 1 (#9441)

2021-01-06 12:17:24 -05:00

dependency_versions_table.py

[Setup.py] update jaxlib (#9831)

2021-01-27 11:34:21 +03:00

file_utils.py

Partial local tokenizer load (#9807)

2021-01-28 03:29:12 -05:00

generation_beam_search.py

Add flags to return scores, hidden states and / or attention weights in GenerationMixin (#9150)

2021-01-06 17:11:42 +01:00

generation_logits_process.py

Delete a needless duplicate condition (#9826)

2021-01-27 13:15:23 +03:00

generation_tf_utils.py

[XLNet] Fix mems behavior (#8567)

2020-11-25 16:54:59 -05:00

generation_utils.py

Update past_key_values in GPT-2 (#9596)

2021-01-19 16:00:15 +01:00

hf_api.py

[hf_api] Fix incorrect typing

2020-12-22 19:52:47 +01:00

hf_argparser.py

[traner] fix --lr_scheduler_type choices (#9800)

2021-01-27 10:12:15 -05:00

integrations.py

Deprecate model_path in Trainer.train (#9854)

2021-01-28 08:32:46 -05:00

modelcard.py

Tokenizers: ability to load from model subfolder (#8586)

2020-11-17 08:58:45 -05:00

modeling_flax_utils.py

Support for private models from huggingface.co (#9141)

2020-12-16 10:09:57 -05:00

modeling_outputs.py

[Docs] past_key_values return a tuple of tuple as a default (#9381)

2021-01-02 15:55:07 +01:00

modeling_tf_outputs.py

Copyright (#8970)

2020-12-07 18:36:34 -05:00

modeling_tf_pytorch_utils.py

ConvBERT Model (#9717)

2021-01-27 03:20:09 -05:00

modeling_tf_utils.py

Clean TF Bert (#9788)

2021-01-27 11:28:11 +01:00

modeling_utils.py

[trainer] --model_parallel hasn't been implemented for most models (#9347)

2021-01-05 04:01:30 -05:00

optimization_tf.py

Copyright (#8970)

2020-12-07 18:36:34 -05:00

optimization.py

Seq2seq trainer (#9241)

2020-12-22 11:33:44 -05:00

testing_utils.py

Fast transformers import part 1 (#9441)

2021-01-06 12:17:24 -05:00

tokenization_utils_base.py

Partial local tokenizer load (#9807)

2021-01-28 03:29:12 -05:00

tokenization_utils_fast.py

Warning about too long input for fast tokenizers too (#8799)

2020-12-02 10:18:28 -05:00

tokenization_utils.py

Refactor prepare_seq2seq_batch (#9524)

2021-01-12 18:19:38 -05:00

trainer_callback.py

feat(wandb): save model as artifact (#8119)

2021-01-05 03:30:46 -05:00

trainer_pt_utils.py

Smdistributed trainer (#9798)

2021-01-26 10:28:21 -05:00

trainer_seq2seq.py

push (#9320)

2020-12-27 21:57:50 +01:00

trainer_tf.py

Fix TFTrainer prediction output (#9662)

2021-01-25 10:27:12 +01:00

trainer_utils.py

Fix auto-resume training from checkpoint (#9822)

2021-01-27 03:48:18 -05:00

trainer.py

pin_memory -> dataloader_pin_memory (#9874)

2021-01-28 21:10:46 +01:00

training_args_seq2seq.py

Seq2seq trainer (#9241)

2020-12-22 11:33:44 -05:00

training_args_tf.py

Add tpu_zone and gcp_project in training_args_tf.py (#9825)

2021-01-27 08:45:09 -05:00

training_args.py

pin_memory -> dataloader_pin_memory (#9874)

2021-01-28 21:10:46 +01:00