Downloads · 30 days
1
3% of all-time downloads
model_hh_shp2_dpo1
guoyu-zhang/model_hh_shp2_dpo1
model_hh_shp2_dpo1 is a machine learning model from guoyu-zhang. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for peft.
should probably proofread and complete it, then remove this comment. --
All-time downloads
29
Public
Repo size
12.6 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors25.2 MB · 93%
At a glance
- Library
- peft
- Access
- Public
- Created
- Apr 16, 2024
- Updated
- Apr 16, 2024
- SHA
- 3e09ccb7
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
model_hh_shp2_dpo1
This model is a fine-tuned version of meta-llama/Llama-2-7b-chat-hf on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.5468
- Rewards/chosen: -6.9137
- Rewards/rejected: -6.8682
- Rewards/accuracies: 0.5
- Rewards/margins: -0.0455
- Logps/rejected: -315.8322
- Logps/chosen: -297.7029
- Logits/rejected: -1.0826
- Logits/chosen: -1.0343
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: 0.0005
- train_batch_size: 4
- eval_batch_size: 1
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- training_steps: 1000
Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.0732 | 2.67 | 100 | 1.0000 | -3.0844 | -2.9869 | 0.4600 | -0.0975 | -277.0198 | -259.4100 | -0.8881 | -0.8929 |
| 0.0011 | 5.33 | 200 | 1.3488 | -4.4604 | -4.2686 | 0.4600 | -0.1918 | -289.8364 | -273.1703 | -1.1160 | -1.0900 |
| 0.0001 | 8.0 | 300 | 1.5079 | -6.5326 | -6.4800 | 0.5100 | -0.0526 | -311.9508 | -293.8922 | -1.0919 | -1.0442 |
| 0.0001 | 10.67 | 400 | 1.5277 | -6.7102 | -6.6582 | 0.5100 | -0.0520 | -313.7332 | -295.6687 | -1.0878 | -1.0396 |
| 0.0001 | 13.33 | 500 | 1.5353 | -6.8159 | -6.7652 | 0.5100 | -0.0506 | -314.8031 | -296.7249 | -1.0848 | -1.0365 |
| 0.0001 | 16.0 | 600 | 1.5452 | -6.8737 | -6.8239 | 0.5100 | -0.0499 | -315.3895 | -297.3038 | -1.0828 | -1.0346 |
| 0.0001 | 18.67 | 700 | 1.5456 | -6.8995 | -6.8549 | 0.5 | -0.0446 | -315.6996 | -297.5615 | -1.0817 | -1.0337 |
| 0.0001 | 21.33 | 800 | 1.5462 | -6.9136 | -6.8675 | 0.5100 | -0.0461 | -315.8255 | -297.7021 | -1.0819 | -1.0338 |
| 0.0001 | 24.0 | 900 | 1.5494 | -6.9171 | -6.8741 | 0.5100 | -0.0430 | -315.8921 | -297.7376 | -1.0826 | -1.0342 |
| 0.0001 | 26.67 | 1000 | 1.5468 | -6.9137 | -6.8682 | 0.5 | -0.0455 | -315.8322 | -297.7029 | -1.0826 | -1.0343 |
Framework versions
- PEFT 0.10.0
- Transformers 4.39.1
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Library
- peft
- Created
- Apr 16, 2024
- Updated
- Apr 16, 2024