update links from sgnlp to sgnlp-models
Browse files
README.md
CHANGED
|
@@ -126,13 +126,13 @@ from sgnlp.models.span_extraction import (
|
|
| 126 |
|
| 127 |
# Load model
|
| 128 |
config = RecconSpanExtractionConfig.from_pretrained(
|
| 129 |
-
"https://storage.googleapis.com/sgnlp/models/reccon_span_extraction/config.json"
|
| 130 |
)
|
| 131 |
tokenizer = RecconSpanExtractionTokenizer.from_pretrained(
|
| 132 |
"mrm8488/spanbert-finetuned-squadv2"
|
| 133 |
)
|
| 134 |
model = RecconSpanExtractionModel.from_pretrained(
|
| 135 |
-
"https://storage.googleapis.com/sgnlp/models/reccon_span_extraction/pytorch_model.bin",
|
| 136 |
config=config,
|
| 137 |
)
|
| 138 |
preprocessor = RecconSpanExtractionPreprocessor(tokenizer)
|
|
@@ -171,8 +171,8 @@ The train and evaluation datasets were derived from the RECCON dataset. The full
|
|
| 171 |
- **Training Time:** ~3 hours for 12 epochs on a single V100 GPU.
|
| 172 |
|
| 173 |
# Model Parameters
|
| 174 |
-
- **Model Weights:** [link](https://storage.googleapis.com/sgnlp/models/reccon_span_extraction/pytorch_model.bin)
|
| 175 |
-
- **Model Config:** [link](https://storage.googleapis.com/sgnlp/models/reccon_span_extraction/config.json)
|
| 176 |
- **Model Inputs:** Target utterance, emotion in target utterance, evidence utterance and conversational history.
|
| 177 |
- **Model Outputs:** Array of start logits and array of end logits. These 2 arrays can be post processed to detemine the start and end of the causal span.
|
| 178 |
- **Model Size:** ~411MB
|
|
|
|
| 126 |
|
| 127 |
# Load model
|
| 128 |
config = RecconSpanExtractionConfig.from_pretrained(
|
| 129 |
+
"https://storage.googleapis.com/sgnlp-models/models/reccon_span_extraction/config.json"
|
| 130 |
)
|
| 131 |
tokenizer = RecconSpanExtractionTokenizer.from_pretrained(
|
| 132 |
"mrm8488/spanbert-finetuned-squadv2"
|
| 133 |
)
|
| 134 |
model = RecconSpanExtractionModel.from_pretrained(
|
| 135 |
+
"https://storage.googleapis.com/sgnlp-models/models/reccon_span_extraction/pytorch_model.bin",
|
| 136 |
config=config,
|
| 137 |
)
|
| 138 |
preprocessor = RecconSpanExtractionPreprocessor(tokenizer)
|
|
|
|
| 171 |
- **Training Time:** ~3 hours for 12 epochs on a single V100 GPU.
|
| 172 |
|
| 173 |
# Model Parameters
|
| 174 |
+
- **Model Weights:** [link](https://storage.googleapis.com/sgnlp-models/models/reccon_span_extraction/pytorch_model.bin)
|
| 175 |
+
- **Model Config:** [link](https://storage.googleapis.com/sgnlp-models/models/reccon_span_extraction/config.json)
|
| 176 |
- **Model Inputs:** Target utterance, emotion in target utterance, evidence utterance and conversational history.
|
| 177 |
- **Model Outputs:** Array of start logits and array of end logits. These 2 arrays can be post processed to detemine the start and end of the causal span.
|
| 178 |
- **Model Size:** ~411MB
|