File size: 1,863 Bytes
e18f5d1 |
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 |
---
tags:
- generated_from_trainer
model-index:
- name: BitLlama2-jp-127M-optim-3
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. -->
# BitLlama2-jp-127M-optim-3
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 3.3126
## 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: 0.0024
- train_batch_size: 96
- eval_batch_size: 96
- seed: 42
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 1
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 6.8077 | 0.07 | 200 | 4.8407 |
| 4.5572 | 0.15 | 400 | 4.3299 |
| 4.2394 | 0.22 | 600 | 4.1002 |
| 4.0211 | 0.29 | 800 | 3.9227 |
| 3.8617 | 0.37 | 1000 | 3.8109 |
| 3.7662 | 0.44 | 1200 | 3.7284 |
| 3.6699 | 0.51 | 1400 | 3.6362 |
| 3.5966 | 0.58 | 1600 | 3.5681 |
| 3.5261 | 0.66 | 1800 | 3.5094 |
| 3.4696 | 0.73 | 2000 | 3.4567 |
| 3.4162 | 0.8 | 2200 | 3.4072 |
| 3.3665 | 0.88 | 2400 | 3.3600 |
| 3.3162 | 0.95 | 2600 | 3.3126 |
### Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
|