Update README.md
Browse files
README.md
CHANGED
|
@@ -1,22 +1,31 @@
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
-
- niltheory/ExistenceTypes
|
| 4 |
language:
|
| 5 |
-
- en
|
| 6 |
---
|
| 7 |
|
| 8 |
-
# Existence Analysis Model
|
| 9 |
-
## Created for: Compendium Terminum, IP.
|
| 10 |
|
| 11 |
-
|
| 12 |
-
|
| 13 |
|
| 14 |
-
|
| 15 |
-
Increased the data training from 96 to 296.
|
| 16 |
-
Which greatly increased the accuracy scores, albeit there is some edge cases.
|
| 17 |
|
| 18 |
-
Iteration #
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
+
- niltheory/ExistenceTypes
|
| 4 |
language:
|
| 5 |
+
- en
|
| 6 |
---
|
| 7 |
|
| 8 |
+
# Existence Analysis Model (EAM)
|
|
|
|
| 9 |
|
| 10 |
+
**Created for**: Compendium Terminum, IP
|
| 11 |
+
**Base Model**: `bert-large-cased-whole-word-masking`
|
| 12 |
|
| 13 |
+
## Iterative Development
|
|
|
|
|
|
|
| 14 |
|
| 15 |
+
### Iteration #1:
|
| 16 |
+
- **Initial Model**: Utilized `distilBert` for foundational training.
|
| 17 |
+
- **Dataset Size**: 96 entries.
|
| 18 |
+
- **Outcome**: Established baseline for accuracy metrics.
|
| 19 |
|
| 20 |
+
### Iteration #2:
|
| 21 |
+
- **Dataset Expansion**: Increased from 96 to 296 entries.
|
| 22 |
+
- **Performance**: Improved accuracy scores; identified edge cases for refinement.
|
| 23 |
+
|
| 24 |
+
### Iteration #3:
|
| 25 |
+
- **Model Upgrade**: Transitioned to `bert-large-cased-whole-word-masking` from `bert-base-uncased`.
|
| 26 |
+
- **Advancements**: Enhanced contextual sensitivity and accuracy.
|
| 27 |
+
- **Results**: Demonstrated more nuanced understanding and sensitivity in predictions.
|
| 28 |
+
|
| 29 |
+
## Observations
|
| 30 |
+
- Each iteration has contributed to the model's evolving sophistication, leading to improved interpretive performance and accuracy.
|
| 31 |
+
- Continuous evaluation, especially in complex or ambiguous cases, is pivotal for future enhancements.
|