Instructions to use gaussalgo/T5-LM-Large_Canard-HotpotQA-rephrase with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use gaussalgo/T5-LM-Large_Canard-HotpotQA-rephrase with Transformers:
# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("gaussalgo/T5-LM-Large_Canard-HotpotQA-rephrase") model = AutoModelForSeq2SeqLM.from_pretrained("gaussalgo/T5-LM-Large_Canard-HotpotQA-rephrase", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Commit ·
2ae34df
1
Parent(s): 90ff47d
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
- hotpot_qa
|
| 4 |
-
-
|
| 5 |
---
|
| 6 |
|
| 7 |
# Model Card for T5-LM-Large_Canard-HotpotQA-rephrase
|
|
@@ -15,8 +15,8 @@ The best checkpoint was picked by a minimal loss on all (3) training objectives.
|
|
| 15 |
|
| 16 |
```python
|
| 17 |
import datasets
|
| 18 |
-
canard_train_augm = datasets.
|
| 19 |
-
canard_test_augm = datasets.
|
| 20 |
|
| 21 |
canard_df = canard_train_augm.to_pandas()
|
| 22 |
canard_test_df = canard_train_augm.to_pandas()
|
|
@@ -105,6 +105,9 @@ rephrasing_val_labels = canard_test_df.Rewrite.values
|
|
| 105 |
print(rephrasing_labels[0])
|
| 106 |
|
| 107 |
# Training
|
|
|
|
|
|
|
|
|
|
| 108 |
from adaptor.lang_module import LangModule
|
| 109 |
|
| 110 |
lang_module = LangModule("google/t5-large-lm-adapt")
|
|
@@ -191,4 +194,4 @@ See the prompting templates used in training to infer the optimal prompting form
|
|
| 191 |
|
| 192 |
#### Contact
|
| 193 |
|
| 194 |
-
Feel free to ask questions at stefanik{at} gaussalgo.com
|
|
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
- hotpot_qa
|
| 4 |
+
- gaussalgo/Canard_Wiki-augmented
|
| 5 |
---
|
| 6 |
|
| 7 |
# Model Card for T5-LM-Large_Canard-HotpotQA-rephrase
|
|
|
|
| 15 |
|
| 16 |
```python
|
| 17 |
import datasets
|
| 18 |
+
canard_train_augm = datasets.load_dataset("gaussalgo/Canard_Wiki-augmented", split="train") # see the dataset card for details
|
| 19 |
+
canard_test_augm = datasets.load_dataset("gaussalgo/Canard_Wiki-augmented", split="test")
|
| 20 |
|
| 21 |
canard_df = canard_train_augm.to_pandas()
|
| 22 |
canard_test_df = canard_train_augm.to_pandas()
|
|
|
|
| 105 |
print(rephrasing_labels[0])
|
| 106 |
|
| 107 |
# Training
|
| 108 |
+
# see Adaptor's homepage for details:
|
| 109 |
+
# https://github.com/gaussalgo/adaptor
|
| 110 |
+
|
| 111 |
from adaptor.lang_module import LangModule
|
| 112 |
|
| 113 |
lang_module = LangModule("google/t5-large-lm-adapt")
|
|
|
|
| 194 |
|
| 195 |
#### Contact
|
| 196 |
|
| 197 |
+
Feel free to ask questions here, or at stefanik{at} gaussalgo.com
|