Skip to content

rust: support images and generic blob sequences #4564

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wchargin merged 1 commit into from
Jan 19, 2021
Merged

rust: support images and generic blob sequences #4564

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
178 changes: 173 additions & 5 deletions tensorboard/data/server/data_compat.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -23,6 +23,7 @@ use crate::proto::tensorboard as pb;
use pb::summary_metadata::PluginData;

pub(crate) const SCALARS_PLUGIN_NAME: &str = "scalars";
pub(crate) const IMAGES_PLUGIN_NAME: &str = "images";
pub(crate) const GRAPHS_PLUGIN_NAME: &str = "graphs";

/// The inner contents of a single value from an event.
Expand Down Expand Up @@ -70,11 +71,29 @@ impl EventValue {

/// Consumes this event value and enriches it into a blob sequence.
///
/// For now, this succeeds only for graphs.
/// For now, this supports `GraphDef`s, summaries with `image`, or summaries with `tensor` set
/// to a rank-1 tensor of type `DT_STRING`.
pub fn into_blob_sequence(self) -> Result<BlobSequenceValue, DataLoss> {
match self {
EventValue::Summary(_) => Err(DataLoss),
EventValue::GraphDef(GraphDefValue(blob)) => Ok(BlobSequenceValue(vec![blob])),
EventValue::Summary(SummaryValue(value_box)) => match *value_box {
pb::summary::value::Value::Image(im) => {
let w = format!("{}", im.width).into_bytes();
let h = format!("{}", im.height).into_bytes();
let buf = im.encoded_image_string;
Ok(BlobSequenceValue(vec![w, h, buf]))
}
pb::summary::value::Value::Tensor(tp) => {
if tp.dtype == i32::from(pb::DataType::DtString)
&& tp.tensor_shape.map_or(false, |shape| shape.dim.len() == 1)
{
Ok(BlobSequenceValue(tp.string_val))
} else {
Err(DataLoss)
}
}
_ => Err(DataLoss),
},
}
}
}
Expand Down Expand Up @@ -166,13 +185,16 @@ impl SummaryValue {
// form.
(Some(md), _) if md.data_class != i32::from(pb::DataClass::Unknown) => Box::new(md),
(_, Value::SimpleValue(_)) => blank(SCALARS_PLUGIN_NAME, pb::DataClass::Scalar),
(_, Value::Image(_)) => blank(IMAGES_PLUGIN_NAME, pb::DataClass::BlobSequence),
(Some(mut md), _) => {
// Use given metadata, but first set data class based on plugin name, if known.
#[allow(clippy::single_match)] // will have more patterns later
match md.plugin_data.as_ref().map(|pd| pd.plugin_name.as_str()) {
Some(SCALARS_PLUGIN_NAME) => {
md.data_class = pb::DataClass::Scalar.into();
}
Some(IMAGES_PLUGIN_NAME) => {
md.data_class = pb::DataClass::BlobSequence.into();
}
_ => {}
};
Box::new(md)
Expand Down Expand Up @@ -461,16 +483,72 @@ mod tests {
}
}

mod graphs {
mod blob_sequences {
use super::*;

#[test]
fn test_metadata() {
fn test_metadata_graph() {
let md = GraphDefValue::initial_metadata();
assert_eq!(&md.plugin_data.unwrap().plugin_name, GRAPHS_PLUGIN_NAME);
assert_eq!(md.data_class, i32::from(pb::DataClass::BlobSequence));
}

#[test]
fn test_metadata_tf1x_image() {
let v = SummaryValue(Box::new(Value::Image(pb::summary::Image {
height: 480,
width: 640,
colorspace: 3,
encoded_image_string: b"\x89PNGabc".to_vec(),
..Default::default()
})));
let result = v.initial_metadata(None);

assert_eq!(
*result,
pb::SummaryMetadata {
plugin_data: Some(PluginData {
plugin_name: IMAGES_PLUGIN_NAME.to_string(),
..Default::default()
}),
data_class: pb::DataClass::BlobSequence.into(),
..Default::default()
}
);
}

#[test]
fn test_metadata_tf2x_image_without_dataclass() {
let md = pb::SummaryMetadata {
plugin_data: Some(PluginData {
plugin_name: IMAGES_PLUGIN_NAME.to_string(),
content: b"preserved!".to_vec(),
..Default::default()
}),
..Default::default()
};
let v = SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[2])),
string_val: vec![b"\x89PNGabc".to_vec(), b"\x89PNGdef".to_vec()],
..Default::default()
})));
let result = v.initial_metadata(Some(md));

assert_eq!(
*result,
pb::SummaryMetadata {
plugin_data: Some(PluginData {
plugin_name: IMAGES_PLUGIN_NAME.to_string(),
content: b"preserved!".to_vec(),
..Default::default()
}),
data_class: pb::DataClass::BlobSequence.into(),
..Default::default()
}
);
}

#[test]
fn test_enrich_graph_def() {
let v = EventValue::GraphDef(GraphDefValue(vec![1, 2, 3, 4]));
Expand All @@ -479,6 +557,96 @@ mod tests {
Ok(BlobSequenceValue(vec![vec![1, 2, 3, 4]]))
);
}

#[test]
fn test_enrich_tf1x_image() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Image(pb::summary::Image {
height: 480,
width: 640,
colorspace: 3,
encoded_image_string: b"\x89PNGabc".to_vec(),
..Default::default()
}))));
let expected = BlobSequenceValue(vec![
b"640".to_vec(),
b"480".to_vec(),
b"\x89PNGabc".to_vec(),
]);
assert_eq!(v.into_blob_sequence(), Ok(expected));
}

#[test]
fn test_enrich_valid_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[2])),
string_val: vec![b"abc".to_vec(), b"defghi".to_vec()],
..Default::default()
}))));
let expected = BlobSequenceValue(vec![b"abc".to_vec(), b"defghi".to_vec()]);
assert_eq!(v.into_blob_sequence(), Ok(expected));
}

#[test]
fn test_enrich_valid_empty_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[0])),
string_val: vec![],
..Default::default()
}))));
let expected = BlobSequenceValue(vec![]);
assert_eq!(v.into_blob_sequence(), Ok(expected));
}

#[test]
fn test_enrich_invalid_empty_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[0, 3])), // bad rank
string_val: vec![],
..Default::default()
}))));
assert_eq!(v.into_blob_sequence(), Err(DataLoss));
}

#[test]
fn test_enrich_scalar_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[])),
string_val: vec![b"no scalars for you".to_vec()],
..Default::default()
}))));
assert_eq!(v.into_blob_sequence(), Err(DataLoss));
}

#[test]
fn test_enrich_higher_rank_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtString.into(),
tensor_shape: Some(tensor_shape(&[2, 2])),
string_val: vec![
b"ab".to_vec(),
b"cd".to_vec(),
b"ef".to_vec(),
b"gh".to_vec(),
],
..Default::default()
}))));
assert_eq!(v.into_blob_sequence(), Err(DataLoss));
}

#[test]
fn test_enrich_non_string_tensor() {
let v = EventValue::Summary(SummaryValue(Box::new(Value::Tensor(pb::TensorProto {
dtype: pb::DataType::DtFloat.into(),
tensor_shape: Some(tensor_shape(&[2])),
float_val: vec![1.0, 2.0],
..Default::default()
}))));
assert_eq!(v.into_blob_sequence(), Err(DataLoss));
}
}

mod unknown {
Expand Down