Feature Extraction
Transformers.js
ONNX
English
bert
decidekit
system-one
text-classification
zero-shot-classification
calibration
on-device
Instructions to use debuffjp/decide-small with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers.js
How to use debuffjp/decide-small with Transformers.js:
// npm i @huggingface/transformers import { pipeline } from '@huggingface/transformers'; // Allocate pipeline const pipe = await pipeline('feature-extraction', 'debuffjp/decide-small');
Numbers for decidekit 0.2
Browse files
README.md
CHANGED
|
@@ -43,13 +43,13 @@ The files download from this repo on first use, then stay in the browser's cache
|
|
| 43 |
|
| 44 |
| | |
|
| 45 |
|---|---|
|
| 46 |
-
| BANKING77, 77 labels, all training data |
|
| 47 |
-
| CLINC150, 150 labels, all training data |
|
| 48 |
-
| BANKING77 with 5 examples per label |
|
| 49 |
| A decision in Chrome (worker, median) | 7.7 ms |
|
| 50 |
| Download at 8-bit | 34 MB |
|
| 51 |
|
| 52 |
-
Ryzen 7 260 laptop, Chrome 156. Results and scripts: [eval/results-
|
| 53 |
|
| 54 |
## Limits
|
| 55 |
|
|
|
|
| 43 |
|
| 44 |
| | |
|
| 45 |
|---|---|
|
| 46 |
+
| BANKING77, 77 labels, all training data | 93.1% |
|
| 47 |
+
| CLINC150, 150 labels, all training data | 96.3% |
|
| 48 |
+
| BANKING77 with 5 examples per label | 83.7% |
|
| 49 |
| A decision in Chrome (worker, median) | 7.7 ms |
|
| 50 |
| Download at 8-bit | 34 MB |
|
| 51 |
|
| 52 |
+
Ryzen 7 260 laptop, Chrome 156. Results and scripts: [eval/results-h14.json](https://github.com/debuffjp/decidekit/blob/main/eval/results-h14.json).
|
| 53 |
|
| 54 |
## Limits
|
| 55 |
|