Hugging Face's logo

HuggingFaceTB
/
SmolLM-1.7B-Instruct

Text Generation
Transformers
ONNX
Safetensors
English
llama
alignment-handbook
trl
sft
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
15
SmolLM-1.7B-Instruct / README.md
loubnabnl's picture
loubnabnl HF staff
End of training
6e74f71 verified
preview code
|
raw
history blame
1.93 kB
metadata
base_model: HuggingFaceTB/cosmo2-1.7B-webinst-sc2
tags:
  - alignment-handbook
  - trl
  - dpo
  - generated_from_trainer
  - trl
  - dpo
  - generated_from_trainer
datasets:
  - HuggingFaceTB/Helpsteer
model-index:
  - name: cosmo2-1.7B-webinst-sc2-dpo-helpsteer-ep1
    results: []

Visualize in Weights & Biases

cosmo2-1.7B-webinst-sc2-dpo-helpsteer-ep1

This model is a fine-tuned version of HuggingFaceTB/cosmo2-1.7B-webinst-sc2 on the HuggingFaceTB/Helpsteer dataset. It achieves the following results on the evaluation set:

  • Loss: 0.6672
  • Rewards/chosen: -0.0466
  • Rewards/rejected: -0.0933
  • Rewards/accuracies: 0.5500
  • Rewards/margins: 0.0467
  • Logps/rejected: -149.4311
  • Logps/chosen: -121.9851
  • Logits/rejected: 0.8632
  • Logits/chosen: 0.9551

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-06
  • train_batch_size: 2
  • eval_batch_size: 4
  • seed: 42
  • distributed_type: multi-GPU
  • num_devices: 8
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 128
  • total_eval_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: 1

Training results

Framework versions

  • Transformers 4.42.3
  • Pytorch 2.1.2
  • Datasets 2.20.0
  • Tokenizers 0.19.1