File size: 4,816 Bytes
a56bc69 e42bbb8 a84e55a a56bc69 e42bbb8 a84e55a e42bbb8 a56bc69 e42bbb8 a56bc69 e42bbb8 a56bc69 e42bbb8 a56bc69 e42bbb8 a56bc69 e42bbb8 a56bc69 e42bbb8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 |
---
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
model-index:
- name: tiny-llama
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# tiny-llama
This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v1.0](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v1.0) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6966
- Accuracy: 0.8195
- Precision: 0.8222
- Recall: 0.8195
- Precision Macro: 0.7955
- Recall Macro: 0.7536
- Macro Fpr: 0.0148
- Weighted Fpr: 0.0141
- Weighted Specificity: 0.9765
- Macro Specificity: 0.9873
- Weighted Sensitivity: 0.8327
- Macro Sensitivity: 0.7536
- F1 Micro: 0.8327
- F1 Macro: 0.7609
- F1 Weighted: 0.8291
## 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: 5e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | Precision Macro | Recall Macro | Macro Fpr | Weighted Fpr | Weighted Specificity | Macro Specificity | Weighted Sensitivity | Macro Sensitivity | F1 Micro | F1 Macro | F1 Weighted |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:---------------:|:------------:|:---------:|:------------:|:--------------------:|:-----------------:|:--------------------:|:-----------------:|:--------:|:--------:|:-----------:|
| 1.0444 | 1.0 | 642 | 0.5968 | 0.8056 | 0.8050 | 0.8056 | 0.7122 | 0.6995 | 0.0175 | 0.0169 | 0.9730 | 0.9852 | 0.8056 | 0.6995 | 0.8056 | 0.6986 | 0.8014 |
| 0.4788 | 2.0 | 1284 | 0.6966 | 0.8195 | 0.8222 | 0.8195 | 0.8092 | 0.7825 | 0.0161 | 0.0155 | 0.9755 | 0.9863 | 0.8195 | 0.7825 | 0.8195 | 0.7849 | 0.8172 |
| 0.3354 | 3.0 | 1926 | 0.8046 | 0.8327 | 0.8276 | 0.8327 | 0.8058 | 0.7582 | 0.0148 | 0.0141 | 0.9758 | 0.9872 | 0.8327 | 0.7582 | 0.8327 | 0.7742 | 0.8282 |
| 0.0571 | 4.0 | 2569 | 1.1143 | 0.8265 | 0.8312 | 0.8265 | 0.7904 | 0.7763 | 0.0152 | 0.0148 | 0.9772 | 0.9869 | 0.8265 | 0.7763 | 0.8265 | 0.7690 | 0.8262 |
| 0.0187 | 5.0 | 3211 | 1.1104 | 0.8319 | 0.8316 | 0.8319 | 0.7745 | 0.7724 | 0.0149 | 0.0142 | 0.9770 | 0.9873 | 0.8319 | 0.7724 | 0.8319 | 0.7638 | 0.8303 |
| 0.0071 | 6.0 | 3853 | 1.1445 | 0.8242 | 0.8210 | 0.8242 | 0.7684 | 0.7384 | 0.0157 | 0.0150 | 0.9755 | 0.9866 | 0.8242 | 0.7384 | 0.8242 | 0.7451 | 0.8209 |
| 0.0002 | 7.0 | 4495 | 1.2032 | 0.8327 | 0.8302 | 0.8327 | 0.7985 | 0.7529 | 0.0148 | 0.0141 | 0.9765 | 0.9873 | 0.8327 | 0.7529 | 0.8327 | 0.7617 | 0.8293 |
| 0.0028 | 8.0 | 5138 | 1.1918 | 0.8257 | 0.8226 | 0.8257 | 0.7738 | 0.7493 | 0.0155 | 0.0149 | 0.9756 | 0.9868 | 0.8257 | 0.7493 | 0.8257 | 0.7552 | 0.8229 |
| 0.0 | 9.0 | 5780 | 1.2181 | 0.8311 | 0.8286 | 0.8311 | 0.7935 | 0.7522 | 0.0150 | 0.0143 | 0.9764 | 0.9872 | 0.8311 | 0.7522 | 0.8311 | 0.7592 | 0.8276 |
| 0.0018 | 10.0 | 6420 | 1.2265 | 0.8327 | 0.8301 | 0.8327 | 0.7955 | 0.7536 | 0.0148 | 0.0141 | 0.9765 | 0.9873 | 0.8327 | 0.7536 | 0.8327 | 0.7609 | 0.8291 |
### Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.1
|