| --- |
| language: |
| - hi |
| - en |
| library_name: transformers |
| tags: |
| - sentiment-analysis |
| - hinglish |
| - onnx |
| --- |
| |
| # Hinglish sentiment models |
|
|
| Two Hinglish (Hindi-English code-mixed) sentiment models. |
|
|
| | Folder | Description | |
| |---|---| |
| | `model_1.0/` | Newer model with better accuracy. Contains `hinglish_model_checkpoint` (PyTorch), `hinglish_onnx_model` (ONNX) and `hinglish_onnx_fp16` (optimized fp16 ONNX). | |
| | `model_-1.0/` | Older model with better speed (optimized ONNX). | |
|
|
| ## Usage |
|
|
| ```python |
| from transformers import AutoTokenizer, AutoModelForSequenceClassification |
| |
| repo = "v-krishna07/hinglish-models" |
| sub = "model_1.0/hinglish_model_checkpoint" |
| tok = AutoTokenizer.from_pretrained(repo, subfolder=sub) |
| model = AutoModelForSequenceClassification.from_pretrained(repo, subfolder=sub) |
| ``` |
|
|
| For the ONNX folders, use `optimum`: |
|
|
| ```python |
| from optimum.onnxruntime import ORTModelForSequenceClassification |
| |
| model = ORTModelForSequenceClassification.from_pretrained( |
| repo, subfolder="model_-1.0", file_name="model_optimized.onnx" |
| ) |
| ``` |
|
|
