SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

finish updating docstrings

Browse Source
This commit is contained in:
Joel Grus
2019-02-23 06:31:59 -08:00
parent 33aa7a80ca
commit 8722e9eb3b
2 changed files with 21 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
README.md
View File

@@ -773,7 +773,7 @@ This model *outputs*:
*Outputs*: *Outputs*:
- if `lm_labels` is not `None`: - if `lm_labels` is not `None`:
Outputs the language modeling loss. Outputs the language modeling loss.
- else: a tupple of - else: a tuple of
- `lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, sequence_length, total_tokens_embeddings] (or more generally [d_1, ..., d_n, total_tokens_embeddings] were d_1 ... d_n are the dimension of input_ids) - `lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, sequence_length, total_tokens_embeddings] (or more generally [d_1, ..., d_n, total_tokens_embeddings] were d_1 ... d_n are the dimension of input_ids)
- `presents`: a list of pre-computed hidden-states (key and values in each attention blocks) as a torch.FloatTensors. They can be reused to speed up sequential decoding (see the `run_gpt2.py` example). - `presents`: a list of pre-computed hidden-states (key and values in each attention blocks) as a torch.FloatTensors. They can be reused to speed up sequential decoding (see the `run_gpt2.py` example).

18
pytorch_pretrained_bert/modeling_gpt2.py
View File

@@ -492,12 +492,16 @@ class GPT2Model(GPT2PreTrainedModel):
(the previous two being the word and position embeddings). (the previous two being the word and position embeddings).
The input, position and token_type embeddings are summed inside the Transformer before the first The input, position and token_type embeddings are summed inside the Transformer before the first
self-attention block. self-attention block.
`past`: an optional list of torch.LongTensor that contains pre-computed hidden-states
(key and values in the attention blocks) to speed up sequential decoding
(this is the presents output of the model, cf. below).
Outputs a tuple consisting of: Outputs a tuple consisting of:
`hidden_states`: the encoded-hidden-states at the top of the model `hidden_states`: the encoded-hidden-states at the top of the model
as a torch.FloatTensor of size [batch_size, sequence_length, hidden_size] as a torch.FloatTensor of size [batch_size, sequence_length, hidden_size]
(or more generally [d_1, ..., d_n, hidden_size] were d_1 ... d_n are the dimension of input_ids) (or more generally [d_1, ..., d_n, hidden_size] were d_1 ... d_n are the dimension of input_ids)
`presents`: ? `presents`: a list of pre-computed hidden-states (key and values in each attention blocks) as
torch.FloatTensors. They can be reused to speed up sequential decoding.
Example usage: Example usage:
```python ```python
@@ -571,6 +575,9 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
`lm_labels`: optional language modeling labels: torch.LongTensor of shape [batch_size, sequence_length] `lm_labels`: optional language modeling labels: torch.LongTensor of shape [batch_size, sequence_length]
with indices selected in [-1, 0, ..., vocab_size]. All labels set to -1 are ignored (masked), the loss with indices selected in [-1, 0, ..., vocab_size]. All labels set to -1 are ignored (masked), the loss
is only computed for the labels set in [0, ..., vocab_size] is only computed for the labels set in [0, ..., vocab_size]
`past`: an optional list of torch.LongTensor that contains pre-computed hidden-states
(key and values in the attention blocks) to speed up sequential decoding
(this is the presents output of the model, cf. below).
Outputs: Outputs:
if `lm_labels` is not `None`: if `lm_labels` is not `None`:
@@ -578,7 +585,8 @@ class GPT2LMHeadModel(GPT2PreTrainedModel):
else a tuple: else a tuple:
`lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, sequence_length, config.vocab_size] `lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, sequence_length, config.vocab_size]
(or more generally [d_1, ..., d_n, config.vocab_size] were d_1 ... d_n are the dimension of input_ids) (or more generally [d_1, ..., d_n, config.vocab_size] were d_1 ... d_n are the dimension of input_ids)
`presents`: ... `presents`: a list of pre-computed hidden-states (key and values in each attention blocks) as
torch.FloatTensors. They can be reused to speed up sequential decoding.
Example usage: Example usage:
```python ```python
@@ -636,6 +644,9 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
is only computed for the labels set in [0, ..., config.vocab_size] is only computed for the labels set in [0, ..., config.vocab_size]
`multiple_choice_labels`: optional multiple choice labels: torch.LongTensor of shape [batch_size] `multiple_choice_labels`: optional multiple choice labels: torch.LongTensor of shape [batch_size]
with indices selected in [0, ..., num_choices]. with indices selected in [0, ..., num_choices].
`past`: an optional list of torch.LongTensor that contains pre-computed hidden-states
(key and values in the attention blocks) to speed up sequential decoding
(this is the presents output of the model, cf. below).
Outputs: Outputs:
if `lm_labels` and `multiple_choice_labels` are not `None`: if `lm_labels` and `multiple_choice_labels` are not `None`:
@@ -643,7 +654,8 @@ class GPT2DoubleHeadsModel(GPT2PreTrainedModel):
else: a tuple with else: a tuple with
`lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, num_choices, sequence_length, config.vocab_size] `lm_logits`: the language modeling logits as a torch.FloatTensor of size [batch_size, num_choices, sequence_length, config.vocab_size]
`multiple_choice_logits`: the multiple choice logits as a torch.FloatTensor of size [batch_size, num_choices] `multiple_choice_logits`: the multiple choice logits as a torch.FloatTensor of size [batch_size, num_choices]
`presents`: ... `presents`: a list of pre-computed hidden-states (key and values in each attention blocks) as
torch.FloatTensors. They can be reused to speed up sequential decoding.
Example usage: Example usage:
```python ```python