Artigenz-Coder-DS-6.7B_components_dataset_size_52_epochs_10_2024-06-12_23-06-20_3525894
This model is a fine-tuned version of Artigenz/Artigenz-Coder-DS-6.7B on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2645
- Accuracy: 0.472
- Chrf: 0.857
- Bleu: 0.787
- Sacrebleu: 0.8
- Rouge1: 0.854
- Rouge2: 0.749
- Rougel: 0.836
- Rougelsum: 0.848
- Meteor: 0.856
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: 0.001
- train_batch_size: 1
- eval_batch_size: 1
- seed: 3407
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 4
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-06
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 52
- training_steps: 520
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Chrf | Bleu | Sacrebleu | Rouge1 | Rouge2 | Rougel | Rougelsum | Meteor |
---|---|---|---|---|---|---|---|---|---|---|---|---|
0.0027 | 0.83 | 52 | 0.3670 | 0.472 | 0.817 | 0.726 | 0.7 | 0.83 | 0.708 | 0.805 | 0.824 | 0.803 |
0.0158 | 1.65 | 104 | 0.3329 | 0.472 | 0.83 | 0.746 | 0.7 | 0.831 | 0.708 | 0.809 | 0.824 | 0.825 |
0.0281 | 2.48 | 156 | 0.2940 | 0.471 | 0.84 | 0.758 | 0.8 | 0.837 | 0.719 | 0.817 | 0.831 | 0.828 |
0.0095 | 3.3 | 208 | 0.3632 | 0.472 | 0.814 | 0.728 | 0.7 | 0.818 | 0.673 | 0.791 | 0.811 | 0.796 |
0.2691 | 4.13 | 260 | 0.2944 | 0.472 | 0.844 | 0.768 | 0.8 | 0.844 | 0.724 | 0.822 | 0.838 | 0.85 |
0.0012 | 4.95 | 312 | 0.2783 | 0.472 | 0.852 | 0.777 | 0.8 | 0.85 | 0.738 | 0.828 | 0.844 | 0.852 |
0.0155 | 5.78 | 364 | 0.2727 | 0.472 | 0.856 | 0.783 | 0.8 | 0.849 | 0.738 | 0.829 | 0.844 | 0.852 |
0.0407 | 6.6 | 416 | 0.2687 | 0.472 | 0.855 | 0.783 | 0.8 | 0.845 | 0.73 | 0.825 | 0.838 | 0.856 |
0.0139 | 7.43 | 468 | 0.2658 | 0.472 | 0.855 | 0.783 | 0.8 | 0.847 | 0.731 | 0.827 | 0.84 | 0.855 |
0.0008 | 8.25 | 520 | 0.2645 | 0.472 | 0.857 | 0.787 | 0.8 | 0.854 | 0.749 | 0.836 | 0.848 | 0.856 |
Framework versions
- PEFT 0.7.1
- Transformers 4.37.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.2
- Tokenizers 0.15.2
- Downloads last month
- 0
Inference Providers
NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API:
The model has no pipeline_tag.
Model tree for vdavidr/Artigenz-Coder-DS-6.7B_components_dataset_size_52_epochs_10_2024-06-12_23-06-20_3525894
Base model
Artigenz/Artigenz-Coder-DS-6.7B