79 lines
2.7 KiB
Markdown
79 lines
2.7 KiB
Markdown
|
---
|
||
|
|
library_name: transformers
|
||
|
|
base_model: W-61/llama-3-8b-base-sft-ultrachat-8xh200
|
||
|
|
tags:
|
||
|
|
- alignment-handbook
|
||
|
|
- slic-hf
|
||
|
|
- generated_from_trainer
|
||
|
|
datasets:
|
||
|
|
- HuggingFaceH4/ultrafeedback_binarized
|
||
|
|
model-index:
|
||
|
|
- name: llama-3-8b-base-slic-hf-ultrafeedback-8xh200
|
||
|
|
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. -->
|
||
|
|
|
||
|
|
# llama-3-8b-base-slic-hf-ultrafeedback-8xh200
|
||
|
|
|
||
|
|
This model is a fine-tuned version of [W-61/llama-3-8b-base-sft-ultrachat-8xh200](https://huggingface.co/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: 341.9101
|
||
|
|
- Rewards/chosen: -260.8732
|
||
|
|
- Rewards/rejected: -246.9854
|
||
|
|
- Rewards/accuracies: 0.4919
|
||
|
|
- Rewards/margins: -13.8878
|
||
|
|
- Logps/chosen: -260.8732
|
||
|
|
- Logps/rejected: -246.9854
|
||
|
|
- Slic/rank Loss: 81.0369
|
||
|
|
- Slic/ce Loss: 260.8732
|
||
|
|
- Logits/chosen: -0.5998
|
||
|
|
- Logits/rejected: -0.6056
|
||
|
|
|
||
|
|
## 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: 8
|
||
|
|
- gradient_accumulation_steps: 4
|
||
|
|
- total_train_batch_size: 128
|
||
|
|
- total_eval_batch_size: 32
|
||
|
|
- 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 | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/chosen | Logps/rejected | Slic/rank Loss | Slic/ce Loss | Logits/chosen | Logits/rejected |
|
||
|
|
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:------------:|:--------------:|:--------------:|:------------:|:-------------:|:---------------:|
|
||
|
|
| 1368.2239 | 0.4188 | 200 | 345.5856 | -262.1702 | -246.1536 | 0.4874 | -16.0166 | -262.1702 | -246.1536 | 83.4155 | 262.1702 | -0.6125 | -0.6205 |
|
||
|
|
| 1395.7896 | 0.8377 | 400 | 341.9101 | -260.8732 | -246.9854 | 0.4919 | -13.8878 | -260.8732 | -246.9854 | 81.0369 | 260.8732 | -0.5998 | -0.6056 |
|
||
|
|
|
||
|
|
|
||
|
|
### Framework versions
|
||
|
|
|
||
|
|
- Transformers 4.51.0
|
||
|
|
- Pytorch 2.3.1+cu121
|
||
|
|
- Datasets 2.21.0
|
||
|
|
- Tokenizers 0.21.4
|