File size: 2,314 Bytes
580698d 57216ab 580698d 57216ab 580698d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 |
---
library_name: transformers
license: llama3.1
base_model: mlfoundations-dev/evol_tt_5s
tags:
- llama-factory
- full
- trl
- dpo
- llama-factory
- generated_from_trainer
model-index:
- name: simpo-evol_tt_5s
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# simpo-evol_tt_5s
This model is a fine-tuned version of [mlfoundations-dev/evol_tt_5s](https://huggingface.co/mlfoundations-dev/evol_tt_5s) on the mlfoundations-dev/gemma2-ultrafeedback-armorm dataset.
It achieves the following results on the evaluation set:
- Loss: 2.4559
- Rewards/chosen: -44.7747
- Rewards/rejected: -52.1235
- Rewards/accuracies: 0.8001
- Rewards/margins: 7.3488
- Logps/chosen: -4.4775
- Logps/rejected: -5.2123
- Logits/chosen: -0.9505
- Logits/rejected: -0.9573
## 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: 8e-07
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- 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: 1.0
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/chosen | Logps/rejected | Logits/chosen | Logits/rejected |
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:------------:|:--------------:|:-------------:|:---------------:|
| 2.838 | 0.9997 | 442 | 2.4559 | -44.7747 | -52.1235 | 0.8001 | 7.3488 | -4.4775 | -5.2123 | -0.9505 | -0.9573 |
### Framework versions
- Transformers 4.46.1
- Pytorch 2.3.0
- Datasets 3.1.0
- Tokenizers 0.20.3
|