conshare-7b-sft-full
This model is a fine-tuned version of /public/home/qianlima/scow/Llama-2-7b-chat-hf on the data/my dataset. It achieves the following results on the evaluation set:
- Loss: 0.7356
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: 1e-05
- trainbatchsize: 8
- evalbatchsize: 8
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradientaccumulationsteps: 4
- totaltrainbatch_size: 256
- totalevalbatch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lrschedulertype: cosine
- lrschedulerwarmup_ratio: 0.1
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.894 | 0.9973 | 183 | 0.8379 |
| 0.7981 | 2.0 | 367 | 0.7538 |
| 0.7032 | 2.9918 | 549 | 0.7356 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.1.2+cu118
- Datasets 2.20.0
- Tokenizers 0.19.1
评论