metadata
library_name: peft
base_model: Anwaarma/edos_taskB_llama3b_merged2_FINAL
tags:
- base_model:adapter:Anwaarma/edos_taskB_llama3b_merged2_FINAL
- lora
- transformers
metrics:
- accuracy
model-index:
- name: new2
results: []
new2
This model is a fine-tuned version of Anwaarma/edos_taskB_llama3b_merged2_FINAL on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.5177
- Accuracy: 0.5546
- F1 Macro: 0.5101
- F1 Micro: 0.5546
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.PAGED_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_ratio: 0.06
- num_epochs: 40
- label_smoothing_factor: 0.1
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro | F1 Micro |
|---|---|---|---|---|---|---|
| 1.3186 | 1.8598 | 100 | 1.4746 | 0.5802 | 0.5325 | 0.5802 |
| 1.1161 | 3.7103 | 200 | 1.4088 | 0.5947 | 0.5390 | 0.5947 |
| 0.9134 | 5.5607 | 300 | 1.4204 | 0.5638 | 0.4999 | 0.5638 |
| 0.7884 | 7.4112 | 400 | 1.3747 | 0.5638 | 0.5132 | 0.5638 |
| 0.7428 | 9.2617 | 500 | 1.3314 | 0.5638 | 0.5076 | 0.5638 |
Framework versions
- PEFT 0.17.1
- Transformers 4.56.2
- Pytorch 2.8.0+cu126
- Datasets 4.1.1
- Tokenizers 0.22.0