File size: 3,673 Bytes
c78b0b4 |
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 89 90 91 92 93 94 95 96 97 98 99 100 |
---
license: gemma
library_name: peft
tags:
- trl
- reward-trainer
- generated_from_trainer
metrics:
- accuracy
base_model: google/gemma-2b
model-index:
- name: RM-HH-Gemma_harmless_gpt3_20000_gemma2b_shuffleFalse_extractchosenFalse
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. -->
# RM-HH-Gemma_harmless_gpt3_20000_gemma2b_shuffleFalse_extractchosenFalse
This model is a fine-tuned version of [google/gemma-2b](https://huggingface.co/google/gemma-2b) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0236
- Accuracy: 0.9907
## 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: 1.41e-05
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 1.0
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.5999 | 0.03 | 250 | 0.1410 | 0.9717 |
| 0.4197 | 0.06 | 500 | 0.0394 | 0.9882 |
| 0.4246 | 0.08 | 750 | 0.0347 | 0.9895 |
| 0.4616 | 0.11 | 1000 | 0.0345 | 0.9885 |
| 0.4141 | 0.14 | 1250 | 0.0308 | 0.9900 |
| 0.3989 | 0.17 | 1500 | 0.0311 | 0.9887 |
| 0.4122 | 0.19 | 1750 | 0.0299 | 0.9895 |
| 0.4106 | 0.22 | 2000 | 0.0298 | 0.9892 |
| 0.4657 | 0.25 | 2250 | 0.0270 | 0.9905 |
| 0.4311 | 0.28 | 2500 | 0.0304 | 0.9890 |
| 0.4474 | 0.31 | 2750 | 0.0277 | 0.9905 |
| 0.4202 | 0.33 | 3000 | 0.0293 | 0.9892 |
| 0.4487 | 0.36 | 3250 | 0.0287 | 0.9902 |
| 0.4219 | 0.39 | 3500 | 0.0257 | 0.9910 |
| 0.4525 | 0.42 | 3750 | 0.0264 | 0.9910 |
| 0.3805 | 0.44 | 4000 | 0.0277 | 0.9897 |
| 0.3824 | 0.47 | 4250 | 0.0241 | 0.9910 |
| 0.4217 | 0.5 | 4500 | 0.0235 | 0.9912 |
| 0.4275 | 0.53 | 4750 | 0.0259 | 0.9905 |
| 0.4395 | 0.56 | 5000 | 0.0247 | 0.9910 |
| 0.3848 | 0.58 | 5250 | 0.0250 | 0.9910 |
| 0.4297 | 0.61 | 5500 | 0.0249 | 0.9900 |
| 0.4167 | 0.64 | 5750 | 0.0258 | 0.9892 |
| 0.4205 | 0.67 | 6000 | 0.0244 | 0.9902 |
| 0.4072 | 0.69 | 6250 | 0.0264 | 0.9890 |
| 0.4033 | 0.72 | 6500 | 0.0253 | 0.9892 |
| 0.3699 | 0.75 | 6750 | 0.0244 | 0.9905 |
| 0.4101 | 0.78 | 7000 | 0.0259 | 0.9887 |
| 0.3969 | 0.81 | 7250 | 0.0249 | 0.9892 |
| 0.3845 | 0.83 | 7500 | 0.0236 | 0.9907 |
| 0.4208 | 0.86 | 7750 | 0.0232 | 0.9907 |
| 0.3925 | 0.89 | 8000 | 0.0232 | 0.9907 |
| 0.3769 | 0.92 | 8250 | 0.0231 | 0.9912 |
| 0.4323 | 0.94 | 8500 | 0.0232 | 0.9912 |
| 0.3999 | 0.97 | 8750 | 0.0236 | 0.9907 |
### Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2 |