File size: 4,906 Bytes
3829c7f 21a5b07 3829c7f 21a5b07 3829c7f 21a5b07 3829c7f 21a5b07 3829c7f 4d259a0 3829c7f 9d33454 3829c7f 9d33454 3829c7f 9d33454 3829c7f 9d33454 3829c7f |
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 |
---
library_name: peft
tags:
- alignment-handbook
- trl
- dpo
- generated_from_trainer
base_model: NbAiLab/nb-gpt-j-6B-v2
datasets:
- hugodk-sch/aftonposten_title_prefs
model-index:
- name: aftonposten-6b-align-scan
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. -->
# aftonposten-6b-align-scan
This model is a fine-tuned version of [data/ap-gpt-j-6b-sft-qlora-04-08](https://huggingface.co/data/ap-gpt-j-6b-sft-qlora-04-08) on the hugodk-sch/aftonposten_title_prefs dataset.
It achieves the following results on the evaluation set:
- Loss: 0.5772
- Rewards/chosen: 0.0684
- Rewards/rejected: 0.0623
- Rewards/accuracies: 0.5307
- Rewards/margins: 0.0061
- Logps/rejected: -37.4276
- Logps/chosen: -33.9368
- Logits/rejected: -2.2420
- Logits/chosen: -2.2469
## 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-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 4
### Training results
| Training Loss | Epoch | Step | Logits/chosen | Logits/rejected | Logps/chosen | Logps/rejected | Validation Loss | Rewards/accuracies | Rewards/chosen | Rewards/margins | Rewards/rejected |
|:-------------:|:-----:|:----:|:-------------:|:---------------:|:------------:|:--------------:|:---------------:|:------------------:|:--------------:|:---------------:|:----------------:|
| 0.4711 | 0.26 | 100 | -2.2401 | -2.2352 | -34.0113 | -37.4979 | 0.5755 | 0.5195 | 0.0163 | 0.0032 | 0.0131 |
| 0.5061 | 0.52 | 200 | -2.2385 | -2.2337 | -34.0500 | -37.5455 | 0.5877 | 0.4992 | -0.0108 | 0.0094 | -0.0202 |
| 0.3371 | 0.78 | 300 | -2.2371 | -2.2322 | -34.0344 | -37.5353 | 0.5843 | 0.5278 | 0.0001 | 0.0132 | -0.0131 |
| 0.4001 | 1.04 | 400 | 0.6350 | -0.0073 | 0.0033 | 0.4838 | -0.0106 | -37.5120 | -34.0450 | -2.2353 | -2.2402 |
| 0.3401 | 1.3 | 500 | 0.6238 | -0.0135 | -0.0193 | 0.5141 | 0.0058 | -37.5443 | -34.0539 | -2.2353 | -2.2402 |
| 0.433 | 1.56 | 600 | 0.6143 | 0.0129 | 0.0108 | 0.5245 | 0.0021 | -37.5011 | -34.0161 | -2.2421 | -2.2469 |
| 0.3298 | 1.82 | 700 | 0.5790 | 0.0633 | 0.0499 | 0.5195 | 0.0134 | -37.4453 | -33.9442 | -2.2401 | -2.2450 |
| 0.14 | 2.08 | 800 | 0.5904 | 0.0586 | 0.0544 | 0.5162 | 0.0041 | -37.4389 | -33.9509 | -2.2423 | -2.2472 |
| 0.2302 | 2.34 | 900 | 0.5758 | 0.0851 | 0.0740 | 0.5544 | 0.0111 | -37.4109 | -33.9130 | -2.2448 | -2.2497 |
| 0.2296 | 2.6 | 1000 | 0.5750 | 0.0631 | 0.0552 | 0.5075 | 0.0080 | -37.4378 | -33.9444 | -2.2440 | -2.2489 |
| 0.2798 | 2.86 | 1100 | 0.5483 | 0.0729 | 0.0545 | 0.5428 | 0.0184 | -37.4387 | -33.9303 | -2.2419 | -2.2468 |
| 0.1195 | 3.12 | 1200 | 0.5759 | 0.0672 | 0.0613 | 0.5137 | 0.0059 | -37.4291 | -33.9386 | -2.2424 | -2.2473 |
| 0.1371 | 3.38 | 1300 | 0.5592 | 0.0733 | 0.0574 | 0.5494 | 0.0159 | -37.4346 | -33.9299 | -2.2434 | -2.2483 |
| 0.0993 | 3.64 | 1400 | 0.6130 | 0.0546 | 0.0598 | 0.4871 | -0.0053 | -37.4311 | -33.9566 | -2.2422 | -2.2471 |
| 0.18 | 3.9 | 1500 | 0.5566 | 0.0778 | 0.0602 | 0.5050 | 0.0176 | -37.4306 | -33.9234 | -2.2423 | -2.2472 |
### Framework versions
- PEFT 0.10.0
- Transformers 4.39.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1 |