Text Classification
Transformers
Safetensors
roberta
Generated from Trainer
text-embeddings-inference
Instructions to use IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds") model = AutoModelForSequenceClassification.from_pretrained("IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds", device_map="auto") - Notebooks
- Google Colab
- Kaggle
sentiment-analysis-roberta-base-V1.6_ima_ds
This model is a fine-tuned version of FacebookAI/roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.1926
- Accuracy: 0.5844
- Precision Macro: 0.5521
- Recall Macro: 0.5796
- F1 Macro: 0.5575
- Precision Weighted: 0.6060
- Recall Weighted: 0.5844
- F1 Weighted: 0.5896
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 3407
- optimizer: Use OptimizerNames.ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 50
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision Macro | Recall Macro | F1 Macro | Precision Weighted | Recall Weighted | F1 Weighted |
|---|---|---|---|---|---|---|---|---|---|---|
| 1.6421 | 0.4 | 20 | 1.6189 | 0.2796 | 0.0665 | 0.2033 | 0.0990 | 0.0876 | 0.2796 | 0.1330 |
| 1.6176 | 0.8 | 40 | 1.6047 | 0.3552 | 0.1424 | 0.2711 | 0.1863 | 0.1879 | 0.3552 | 0.2453 |
| 1.5261 | 1.2 | 60 | 1.5614 | 0.3955 | 0.4059 | 0.3849 | 0.3373 | 0.4669 | 0.3955 | 0.3549 |
| 1.5492 | 1.6 | 80 | 1.4296 | 0.4358 | 0.4566 | 0.4123 | 0.3854 | 0.5207 | 0.4358 | 0.4266 |
| 1.0851 | 2.0 | 100 | 1.3699 | 0.4685 | 0.4874 | 0.4543 | 0.4266 | 0.5473 | 0.4685 | 0.4621 |
| 1.1487 | 2.4 | 120 | 1.1921 | 0.5491 | 0.5222 | 0.5344 | 0.5065 | 0.5925 | 0.5491 | 0.5486 |
| 1.2638 | 2.8 | 140 | 1.2074 | 0.5642 | 0.5496 | 0.5483 | 0.5117 | 0.6262 | 0.5642 | 0.5538 |
| 1.1191 | 3.2 | 160 | 1.1625 | 0.5441 | 0.5483 | 0.5519 | 0.5229 | 0.6081 | 0.5441 | 0.5614 |
| 0.6616 | 3.6 | 180 | 1.1361 | 0.6045 | 0.5685 | 0.5905 | 0.5706 | 0.6323 | 0.6045 | 0.6120 |
| 1.213 | 4.0 | 200 | 1.1492 | 0.5919 | 0.5518 | 0.5598 | 0.5485 | 0.5965 | 0.5919 | 0.5882 |
| 0.7761 | 4.4 | 220 | 1.1632 | 0.5970 | 0.5556 | 0.5662 | 0.5545 | 0.6102 | 0.5970 | 0.5959 |
| 0.6592 | 4.8 | 240 | 1.1926 | 0.5844 | 0.5521 | 0.5796 | 0.5575 | 0.6060 | 0.5844 | 0.5896 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1
- Downloads last month
- 1
Model tree for IMA-StreamSolve/sentiment-analysis-roberta-base-V1.6_ima_ds
Base model
FacebookAI/roberta-base