k2.fsa¶
-
class
k2.Fsa(s, negate_scores=False)[source]¶ -
__init__(s, negate_scores=False)[source]¶ Create an Fsa from a string.
The given string s consists of lines with the following format:
When it represents an acceptor:
src_state dest_state label score
When it represents a transducer:
src_state dest_state label aux_label score
The line for the final state consists of only one field:
final_state
Note
Fields are separated by space(s), tab(s) or both. The score field is a float, while other fields are integers.
Caution
The first column has to be non-decreasing.
Caution
The final state has the largest state number. There is only one final state. All arcs that are connected to the final state have label -1.
- Parameters
s (
str) – The input string. Refer to the above comment for its format.negate_scores (
bool) – Optional. If true, the string form has the weights as costs, not scores, so we negate as we read.
-
classmethod
from_tensor(tensor, aux_labels=None)[source]¶ Build an Fsa from a tensor with optional aux_labels.
It is useful when loading an Fsa from file.
- Parameters
tensor (
Tensor) –A torch tensor of dtype torch.int32 with 4 columns. Each row represents an arc. Column 0 is the src_state, column 1 the dest_state, column 2 the label, and column 3 the score.
Caution
Scores are floats and their binary pattern is reinterpreted as integers and saved in a tensor of dtype torch.int32.
aux_labels (
Optional[Tensor]) – Optional. If not None, it associates an aux_label with every arc, so it has as many rows as tensor. It is a 1-D tensor of dtype torch.int32.
- Return type
Fsa- Returns
An instance of Fsa.
-
to_str(negate_scores=False)[source]¶ Convert an Fsa to a string.
Note
The returned string can be used to construct an Fsa.
- Parameters
negate_scores (
bool) – Optional. If true, we negate the score during the conversion.- Return type
str- Returns
A string representation of the Fsa.
-
property
arcs[source]¶ Return the arcs of the Fsa.
Caution
The scores are not contained in the returned tensor. Please use the weights property if you need it.
- Return type
Tensor- Returns
A tensor of dtype torch.int32 with 3 columns. Each row represents an arc. Column 0 is the src_state, column 1 the dest_state, and column 2 the label.
-
property
weights[source]¶ Returns the weights of arcs in the Fsa.
- Return type
Tensor- Returns
A 1-D tensor of dtype torch.float32. It has as many rows as arcs.
-
property
aux_labels[source]¶ Return the aux_labels associated with arcs, if any.
- Return type
Optional[Tensor]- Returns
None or a 1-D tensor of dtype torch.int32 if the Fsa is a transducer. It has as many rows as arcs.
-