mirror of
https://github.com/huggingface/candle.git
synced 2025-06-16 10:38:54 +00:00
Fix the cat implementation + more testing.
This commit is contained in:
@ -817,11 +817,34 @@ impl Tensor {
|
||||
shape: args[0].shape().clone(),
|
||||
});
|
||||
}
|
||||
if dim == 0 {
|
||||
Self::cat0(args)
|
||||
} else {
|
||||
// TODO: Avoid these transpositions and have an implementation that works
|
||||
// for dim != 0...
|
||||
let args: Vec<Tensor> = args
|
||||
.iter()
|
||||
.map(|a| a.transpose(0, dim))
|
||||
.collect::<Result<Vec<_>>>()?;
|
||||
let args: Vec<&Tensor> = args.iter().collect();
|
||||
let cat = Self::cat0(&args)?;
|
||||
cat.transpose(0, dim)
|
||||
}
|
||||
}
|
||||
|
||||
pub fn cat0(args: &[&Self]) -> Result<Self> {
|
||||
if args.is_empty() {
|
||||
return Err(Error::OpRequiresAtLeastOneTensor { op: "cat" });
|
||||
}
|
||||
if args.len() == 1 {
|
||||
return Ok(args[0].clone());
|
||||
}
|
||||
let rank = args[0].rank();
|
||||
let device = args[0].device();
|
||||
let dtype = args[0].dtype();
|
||||
let first_dims = args[0].shape().dims();
|
||||
let mut cat_dims = first_dims.to_vec();
|
||||
cat_dims[dim] = 0;
|
||||
cat_dims[0] = 0;
|
||||
let mut offsets = vec![0usize];
|
||||
for (arg_idx, arg) in args.iter().enumerate() {
|
||||
if arg.dtype() != dtype {
|
||||
@ -848,10 +871,10 @@ impl Tensor {
|
||||
.zip(arg.shape().dims().iter())
|
||||
.enumerate()
|
||||
{
|
||||
if dim == dim_idx {
|
||||
cat_dims[dim] += v2;
|
||||
if dim_idx == 0 {
|
||||
cat_dims[0] += v2;
|
||||
}
|
||||
if dim != dim_idx && v1 != v2 {
|
||||
if dim_idx != 0 && v1 != v2 {
|
||||
// TODO: It would probably be good to have a nicer error message here, i.e.
|
||||
// mention the problematic dimension and the values.
|
||||
mismatch = true;
|
||||
@ -859,7 +882,7 @@ impl Tensor {
|
||||
}
|
||||
if mismatch {
|
||||
return Err(Error::ShapeMismatchCat {
|
||||
dim,
|
||||
dim: 0, // TODO: not the appropriate error message
|
||||
first_shape: args[0].shape().clone(),
|
||||
n: arg_idx + 1,
|
||||
nth_shape: arg.shape().clone(),
|
||||
@ -871,7 +894,7 @@ impl Tensor {
|
||||
let shape = Shape::from(cat_dims);
|
||||
let op = if args.iter().any(|arg| arg.track_op()) {
|
||||
let args: Vec<Tensor> = args.iter().map(|&arg| arg.clone()).collect();
|
||||
Some(Op::Cat(args, dim))
|
||||
Some(Op::Cat(args, 0))
|
||||
} else {
|
||||
None
|
||||
};
|
||||
|
@ -210,18 +210,18 @@ fn cat() -> Result<()> {
|
||||
.t()?
|
||||
.to_vec2::<f32>()?,
|
||||
[
|
||||
[3.0, 4.0, 5.0, 5.0, 5.0],
|
||||
[2.0, 1.0, 2.0, 7.0, 8.0],
|
||||
[1.0, 1.0, 5.0, 5.0, 5.0],
|
||||
[7.0, 8.0, 2.0, 1.0, 2.0]
|
||||
[3.0, 1.0, 4.0, 1.0, 5.0],
|
||||
[2.0, 7.0, 1.0, 8.0, 2.0],
|
||||
[5.0, 5.0, 5.0, 5.0, 5.0],
|
||||
[2.0, 7.0, 1.0, 8.0, 2.0]
|
||||
]
|
||||
);
|
||||
// TODO: This is not the expected answer, to be fixed!
|
||||
assert_eq!(
|
||||
Tensor::cat(&[&t1, &t2], 1)?.to_vec2::<f32>()?,
|
||||
[
|
||||
[3.0, 1.0, 4.0, 1.0, 5.0, 2.0, 7.0, 1.0, 8.0, 2.0],
|
||||
[5.0, 5.0, 5.0, 5.0, 5.0, 2.0, 7.0, 1.0, 8.0, 2.0]
|
||||
[3.0, 1.0, 4.0, 1.0, 5.0, 5.0, 5.0, 5.0, 5.0, 5.0],
|
||||
[2.0, 7.0, 1.0, 8.0, 2.0, 2.0, 7.0, 1.0, 8.0, 2.0]
|
||||
]
|
||||
);
|
||||
Ok(())
|
||||
|
Reference in New Issue
Block a user