zephyr-7b-oasst1
This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on the SeniorKabanocci/oasst1-binarized dataset. It achieves the following results on the evaluation set:
- Loss: 0.6547
- Rewards/chosen: -0.2776
- Rewards/rejected: -0.3797
- Rewards/accuracies: 0.5913
- Rewards/margins: 0.1021
- Logps/rejected: -229.8175
- Logps/chosen: -250.2433
- Logits/rejected: -2.9022
- Logits/chosen: -2.9328
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-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Training results
Framework versions
- PEFT 0.10.0
- Transformers 4.42.3
- Pytorch 2.2.0
- Datasets 2.20.0
- Tokenizers 0.19.1
- Downloads last month
- 1
Model tree for jmajkutewicz/zephyr-7b-oasst1
Base model
mistralai/Mistral-7B-v0.1
Finetuned
alignment-handbook/zephyr-7b-sft-full