Text Classification
Transformers
ONNX
Safetensors
bert
Eval Results (legacy)
text-embeddings-inference
Instructions to use AdamCodd/tinybert-sentiment-amazon with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use AdamCodd/tinybert-sentiment-amazon with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="AdamCodd/tinybert-sentiment-amazon")# Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("AdamCodd/tinybert-sentiment-amazon") model = AutoModelForSequenceClassification.from_pretrained("AdamCodd/tinybert-sentiment-amazon") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,10 +1,34 @@
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
- amazon_polarity
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
---
|
| 5 |
# tinybert-sentiment-amazon
|
| 6 |
|
| 7 |
-
This model is a fine-tuned version of [bert-tiny](prajjwal1/bert-tiny) on [amazon-polarity dataset](https://huggingface.co/datasets/amazon_polarity).
|
|
|
|
|
|
|
|
|
|
| 8 |
|
| 9 |
## Model description
|
| 10 |
|
|
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
- amazon_polarity
|
| 4 |
+
base_model: bert-tiny
|
| 5 |
+
model-index:
|
| 6 |
+
- name: tinybert-sentiment-amazon
|
| 7 |
+
results:
|
| 8 |
+
- task:
|
| 9 |
+
type: text-classification
|
| 10 |
+
name: Text Classification
|
| 11 |
+
dataset:
|
| 12 |
+
name: amazon_polarity
|
| 13 |
+
type: sentiment
|
| 14 |
+
args: default
|
| 15 |
+
metrics:
|
| 16 |
+
- type: accuracy
|
| 17 |
+
value: 0.942
|
| 18 |
+
name: Accuracy
|
| 19 |
+
- type: loss
|
| 20 |
+
value: 0.153
|
| 21 |
+
name: Loss
|
| 22 |
+
- type: f1
|
| 23 |
+
value: 0.940
|
| 24 |
+
name: F1
|
| 25 |
---
|
| 26 |
# tinybert-sentiment-amazon
|
| 27 |
|
| 28 |
+
This model is a fine-tuned version of [bert-tiny](prajjwal1/bert-tiny) on [amazon-polarity dataset](https://huggingface.co/datasets/amazon_polarity). It achieves the following results on the evaluation set:
|
| 29 |
+
* Loss: 0.153
|
| 30 |
+
* Accuracy: 0.942
|
| 31 |
+
* F1_score: 0.940
|
| 32 |
|
| 33 |
## Model description
|
| 34 |
|