G2PTL update
Browse files- README.md +5 -1
- htc_loss.py +1 -1
- modeling_G2PTL.py +1 -1
README.md
CHANGED
|
@@ -28,6 +28,7 @@ More detail: https://arxiv.org/abs/2304.01559
|
|
| 28 |
|
| 29 |
## Intended uses & limitations
|
| 30 |
|
|
|
|
| 31 |
This model is designed for decision tasks based on address text, including tasks related to understanding address texts and Spatial-Temporal downstream tasks which rely on address text representation.
|
| 32 |
|
| 33 |
1. Address text understanding tasks
|
|
@@ -36,11 +37,14 @@ This model is designed for decision tasks based on address text, including tasks
|
|
| 36 |
- Geographic Entity Alignment
|
| 37 |
- Address Text Similarity
|
| 38 |
- Address Texy Classification
|
|
|
|
| 39 |
2. Spatial-Temporal downstream tasks:
|
| 40 |
- Estimated Time of Arrival (ETA) Prediction
|
| 41 |
- Pick-up & Delivery Route Prediction.
|
|
|
|
|
|
|
| 42 |
|
| 43 |
-
The model currently only supports Chinese addresses.
|
| 44 |
|
| 45 |
|
| 46 |
## How to use
|
|
|
|
| 28 |
|
| 29 |
## Intended uses & limitations
|
| 30 |
|
| 31 |
+
|
| 32 |
This model is designed for decision tasks based on address text, including tasks related to understanding address texts and Spatial-Temporal downstream tasks which rely on address text representation.
|
| 33 |
|
| 34 |
1. Address text understanding tasks
|
|
|
|
| 37 |
- Geographic Entity Alignment
|
| 38 |
- Address Text Similarity
|
| 39 |
- Address Texy Classification
|
| 40 |
+
- ...
|
| 41 |
2. Spatial-Temporal downstream tasks:
|
| 42 |
- Estimated Time of Arrival (ETA) Prediction
|
| 43 |
- Pick-up & Delivery Route Prediction.
|
| 44 |
+
- Express Volume Prediction
|
| 45 |
+
- ...
|
| 46 |
|
| 47 |
+
The model currently only supports Chinese addresses, and it is an encoder-only model which is not suitable for text generation scenarios such as question answering. If you need to use address text based dialogue capabilities, you can look forward to our second version of G2PTL (v2.0)
|
| 48 |
|
| 49 |
|
| 50 |
## How to use
|
htc_loss.py
CHANGED
|
@@ -11,7 +11,7 @@ import os
|
|
| 11 |
from transformers.utils.hub import cached_file
|
| 12 |
|
| 13 |
resolved_module_file = cached_file(
|
| 14 |
-
'
|
| 15 |
'htc_mask_dict.pkl',
|
| 16 |
)
|
| 17 |
|
|
|
|
| 11 |
from transformers.utils.hub import cached_file
|
| 12 |
|
| 13 |
resolved_module_file = cached_file(
|
| 14 |
+
'Cainiao-AI/G2PTL',
|
| 15 |
'htc_mask_dict.pkl',
|
| 16 |
)
|
| 17 |
|
modeling_G2PTL.py
CHANGED
|
@@ -20,7 +20,7 @@ import numpy as np
|
|
| 20 |
from .htc_loss import HTCLoss
|
| 21 |
from transformers.utils.hub import cached_file
|
| 22 |
remap_code_2_chn_file_path = cached_file(
|
| 23 |
-
'
|
| 24 |
'remap_code_2_chn.pkl',
|
| 25 |
)
|
| 26 |
|
|
|
|
| 20 |
from .htc_loss import HTCLoss
|
| 21 |
from transformers.utils.hub import cached_file
|
| 22 |
remap_code_2_chn_file_path = cached_file(
|
| 23 |
+
'Cainiao-AI/G2PTL',
|
| 24 |
'remap_code_2_chn.pkl',
|
| 25 |
)
|
| 26 |
|