|
--- |
|
license: apache-2.0 |
|
base_model: hZzy/qwen2.5-0.5b-sft-news-IFT |
|
tags: |
|
- alignment-handbook |
|
- ndcg |
|
- trl |
|
- expo |
|
- generated_from_trainer |
|
- trl |
|
- expo |
|
- generated_from_trainer |
|
datasets: |
|
- hZzy/train_pairwise_weighted |
|
model-index: |
|
- name: qwen2.5-0.5b-expo-DPO-ES-0.1 |
|
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. --> |
|
|
|
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/zhiyuzha-university-of-florida/huggingface/runs/hy7b2tcq) |
|
# qwen2.5-0.5b-expo-DPO-ES-0.1 |
|
|
|
This model is a fine-tuned version of [hZzy/qwen2.5-0.5b-sft-news-IFT](https://huggingface.co/hZzy/qwen2.5-0.5b-sft-news-IFT) on the hZzy/train_pairwise_weighted dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6921 |
|
- Logps: -91.0309 |
|
- Logits: -2.1203 |
|
- Objective: 0.6893 |
|
- Dpo Loss: 0.6893 |
|
- Regularize: 0.6893 |
|
- Ranking Simple: 0.5559 |
|
- Ranking Idealized: 0.6030 |
|
- Ranking Idealized Expo: 0.5223 |
|
- Wo Beta: 7.4262 |
|
|
|
## 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: 4 |
|
- seed: 42 |
|
- distributed_type: multi-GPU |
|
- num_devices: 3 |
|
- gradient_accumulation_steps: 12 |
|
- total_train_batch_size: 144 |
|
- total_eval_batch_size: 12 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_ratio: 0.1 |
|
- num_epochs: 5 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Logps | Logits | Objective | Dpo Loss | Regularize | Ranking Simple | Ranking Idealized | Ranking Idealized Expo | Wo Beta | |
|
|:-------------:|:------:|:----:|:---------------:|:---------:|:-------:|:---------:|:--------:|:----------:|:--------------:|:-----------------:|:----------------------:|:-------:| |
|
| 0.6785 | 0.1417 | 50 | 0.6813 | -90.8716 | -1.6022 | 0.6843 | 0.6843 | 0.6843 | 0.5259 | 0.6030 | 0.5223 | 7.8749 | |
|
| 0.618 | 0.2834 | 100 | 0.6733 | -98.8899 | -1.7799 | 0.6766 | 0.6766 | 0.6766 | 0.5399 | 0.6030 | 0.5223 | 7.7840 | |
|
| 0.5667 | 0.4251 | 150 | 0.6866 | -99.1230 | -1.8072 | 0.6829 | 0.6829 | 0.6829 | 0.5409 | 0.6030 | 0.5223 | 7.8533 | |
|
| 0.5214 | 0.5668 | 200 | 0.6901 | -99.5388 | -1.8894 | 0.6904 | 0.6904 | 0.6904 | 0.5445 | 0.6030 | 0.5223 | 7.6995 | |
|
| 0.4922 | 0.7085 | 250 | 0.6976 | -82.7973 | -1.9880 | 0.6916 | 0.6916 | 0.6916 | 0.5476 | 0.6030 | 0.5223 | 7.8790 | |
|
| 0.4535 | 0.8503 | 300 | 0.6921 | -91.0309 | -2.1203 | 0.6893 | 0.6893 | 0.6893 | 0.5559 | 0.6030 | 0.5223 | 7.4262 | |
|
| 0.423 | 0.9920 | 350 | 0.7057 | -88.1615 | -1.9880 | 0.6959 | 0.6959 | 0.6959 | 0.5549 | 0.6030 | 0.5223 | 7.9979 | |
|
| 0.2847 | 1.1337 | 400 | 0.7315 | -101.6926 | -2.0862 | 0.7281 | 0.7281 | 0.7281 | 0.5424 | 0.6030 | 0.5223 | 8.6326 | |
|
| 0.2991 | 1.2754 | 450 | 0.7008 | -92.7942 | -1.8470 | 0.6980 | 0.6980 | 0.6980 | 0.5621 | 0.6030 | 0.5223 | 8.2584 | |
|
| 0.3065 | 1.4171 | 500 | 0.7180 | -96.6747 | -2.0065 | 0.7147 | 0.7147 | 0.7147 | 0.5554 | 0.6030 | 0.5223 | 8.2522 | |
|
| 0.2895 | 1.5588 | 550 | 0.7044 | -104.2469 | -1.8870 | 0.7077 | 0.7077 | 0.7077 | 0.5652 | 0.6030 | 0.5223 | 8.1947 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.42.0 |
|
- Pytorch 2.3.0+cu121 |
|
- Datasets 3.2.0 |
|
- Tokenizers 0.19.1 |
|
|
