metadata
license: mit
library_name: peft
tags:
- alignment-handbook
- trl
- sft
- generated_from_trainer
- trl
- sft
- generated_from_trainer
datasets:
- ChenWu98/skills_red_herring_metaphor_chat
base_model: HuggingFaceH4/zephyr-7b-beta
model-index:
- name: skills_red_herring_metaphor_chat-lora
results: []
skills_red_herring_metaphor_chat-lora
This model is a fine-tuned version of HuggingFaceH4/zephyr-7b-beta on the ChenWu98/skills_red_herring_metaphor_chat dataset. It achieves the following results on the evaluation set:
- Loss: 0.2900
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: 0.0002
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- distributed_type: multi-GPU
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 4.0
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.8767 | 0.96 | 9 | 0.4226 |
| 0.3349 | 1.92 | 18 | 0.3096 |
| 0.2678 | 2.99 | 28 | 0.2913 |
| 0.2538 | 3.84 | 36 | 0.2900 |
Framework versions
- PEFT 0.7.1
- Transformers 4.37.2
- Pytorch 2.1.2+cu121
- Datasets 2.14.6
- Tokenizers 0.15.1