library_name, base_model, tags, datasets, model-index
| library_name | base_model | tags | datasets | model-index | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| transformers | W-61/llama-3-8b-base-sft-ultrachat-8xh200 |
|
|
|
llama-3-8b-base-r-dpo-ultrafeedback-4xh200
This model is a fine-tuned version of W-61/llama-3-8b-base-sft-ultrachat-8xh200 on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:
- Loss: 0.5080
- R Dpo/chosen Len: 291.2620
- R Dpo/rejected Len: 248.3960
- R Dpo/length Delta: 42.8660
- R Dpo/regularization Term: 0.0
- Logps/chosen: -288.0679
- Logps/rejected: -272.9751
- Logps/ref Chosen: -289.1346
- Logps/ref Rejected: -264.7782
- Logits/chosen: -0.7442
- Logits/rejected: -0.7460
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-07
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 8
- total_train_batch_size: 128
- total_eval_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | R Dpo/chosen Len | R Dpo/rejected Len | R Dpo/length Delta | R Dpo/regularization Term | Logps/chosen | Logps/rejected | Logps/ref Chosen | Logps/ref Rejected | Logits/chosen | Logits/rejected |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 4.178 | 0.4188 | 200 | 0.5266 | 291.2620 | 248.3960 | 42.8660 | 0.0 | -287.2393 | -271.3152 | -289.1346 | -264.7782 | -0.7479 | -0.7490 |
| 4.0423 | 0.8377 | 400 | 0.5080 | 291.2620 | 248.3960 | 42.8660 | 0.0 | -288.0679 | -272.9751 | -289.1346 | -264.7782 | -0.7442 | -0.7460 |
Framework versions
- Transformers 4.51.0
- Pytorch 2.3.1+cu121
- Datasets 2.21.0
- Tokenizers 0.21.4