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

Update pipelines.mdx (#15243)

Browse Source 
fix few spelling mistakes
This commit is contained in:
Kamal Raj
2022-01-20 19:16:48 +05:30
committed by GitHub
parent 85ea462c08
commit 08b41b413a
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
docs/source/main_classes/pipelines.mdx
View File

@@ -254,7 +254,7 @@ For users, a rule of thumb is:
## Pipeline chunk batching ## Pipeline chunk batching
`zero-shot-classification` and `question-answering` are slightly specific in the sense, that a single input might yield `zero-shot-classification` and `question-answering` are slightly specific in the sense, that a single input might yield
mutliple forward pass of a model. Under normal circumstances, this would yield issues with `batch_size` argument. multiple forward pass of a model. Under normal circumstances, this would yield issues with `batch_size` argument.
In order to circumvent this issue, both of these pipelines are a bit specific, they are `ChunkPipeline` instead of In order to circumvent this issue, both of these pipelines are a bit specific, they are `ChunkPipeline` instead of
regular `Pipeline`. In short: regular `Pipeline`. In short:
@@ -263,7 +263,7 @@ regular `Pipeline`. In short:
```python ```python
preprocessed = pipe.preprocess(inputs) preprocessed = pipe.preprocess(inputs)
model_outputs = pipe.forward(preprocessed) model_outputs = pipe.forward(preprocessed)
outputs = pipe.postprocess(model_ouputs) outputs = pipe.postprocess(model_outputs)
``` ```
Now becomes: Now becomes:
@@ -274,7 +274,7 @@ all_model_outputs = []
for preprocessed in pipe.preprocess(inputs): for preprocessed in pipe.preprocess(inputs):
model_outputs = pipe.forward(preprocessed) model_outputs = pipe.forward(preprocessed)
all_model_outputs.append(model_outputs) all_model_outputs.append(model_outputs)
outputs = pipe.postprocess(all_model_ouputs) outputs = pipe.postprocess(all_model_outputs)
``` ```
This should be very transparent to your code because the pipelines are used in This should be very transparent to your code because the pipelines are used in
@@ -282,7 +282,7 @@ the same way.
This is a simplified view, since the pipeline can handle automatically the batch to ! Meaning you don't have to care This is a simplified view, since the pipeline can handle automatically the batch to ! Meaning you don't have to care
about how many forward passes you inputs are actually going to trigger, you can optimize the `batch_size` about how many forward passes you inputs are actually going to trigger, you can optimize the `batch_size`
independantly of the inputs. The caveats from the previous section still apply. independently of the inputs. The caveats from the previous section still apply.
## Pipeline custom code ## Pipeline custom code