metadata
base_model: FacebookAI/roberta-large
language: en
license: apache-2.0
model_name: climate-mitigation-classifier
pipeline_tag: text-classification
tags:
- CRS
- OECD CRS
- text-classification
- lora
- transformers
funded_by: DEval - Deutsches Evaluierungsinstitut der Entwicklungszusammenarbeit gGmbH
tasks:
- text-classification
shared_by: DEval - Deutsches Evaluierungsinstitut der Entwicklungszusammenarbeit gGmbH
This model identifies the relevance of CRS projects to climate-change mitigation. It is trained on manually annotated CRS data using the standard Rio Marker classification. Labels 0, 1, and 2 indicate whether a project has no, significant, or primary focus on climate-change mitigation. (RIO Marker)
Evaluation metrics
| precision | recall | f1-score | support | |
|---|---|---|---|---|
| 0 | 0.92 | 0.90 | 0.91 | 311 |
| 1 | 0.53 | 0.66 | 0.59 | 65 |
| 2 | 0.75 | 0.85 | 0.80 | 87 |
| 3 | 0.59 | 0.37 | 0.46 | 51 |
| -- | -- | -- | -- | -- |
| accuracy | 0.81 | 514 | ||
| macro | avg | 0.70 | 0.70 | 0.69 |
| weighted | avg | 0.81 | 0.81 | 0.81 |
Usage
```python
from transformers import AutoModelForSequenceClassification, AutoTokenizer
model = AutoModelForSequenceClassification.from_pretrained("namespace/my-model")
tokenizer = AutoTokenizer.from_pretrained("namespace/my-model")
inputs = tokenizer("hello world", return_tensors="pt")
outputs = model(**inputs)
print(outputs)"
or
from transformers import TextClassificationPipeline
model = TextClassificationPipeline("namespace/my-model")
outputs = model("Hello World!")
print(outputs)"