YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
Mistral_End
This model is a fine-tuned version of /teamspace/studios/this_studio/output/sft_merge on the WordProblems_SFT_LLama_End dataset. It achieves the following results on the evaluation set:
- Loss: 0.0767
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-05
- train_batch_size: 4
- eval_batch_size: 4
- 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: 500
- num_epochs: 2.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.0893 | 0.2339 | 500 | 0.0972 |
| 0.0753 | 0.4679 | 1000 | 0.0879 |
| 0.071 | 0.7018 | 1500 | 0.0841 |
| 0.0678 | 0.9358 | 2000 | 0.0813 |
| 0.0697 | 1.1697 | 2500 | 0.0799 |
| 0.0553 | 1.4037 | 3000 | 0.0783 |
| 0.057 | 1.6376 | 3500 | 0.0772 |
| 0.0599 | 1.8716 | 4000 | 0.0767 |
Framework versions
- PEFT 0.12.0
- Transformers 4.43.4
- Pytorch 2.3.0
- Datasets 2.20.0
- Tokenizers 0.19.1