Downloads · 30 days
7
15% of all-time downloads
SmolLM-toxic-detox-rlhf
MattBou00/SmolLM-toxic-detox-rlhf
SmolLM-toxic-detox-rlhf is a text generation model from MattBou00. Use it when you need the model to write or continue text. It is set up for transformers.
class Config: basemodel: str = "MattBou00/SmolLM-toxic-finetuned" rewardmodel: str = "s-nlp/robertatoxicityclassifier" datasetname: str = "skrishna/toy-toxicity-dataset" datasetsplit: str = "test" promptwords: int = 2…
All-time downloads
48
Public
Parameters
135M
538 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors538 MB · 99%
At a glance
- Task
- Text Generation
- Library
- transformers
- Model type
- llama
- Access
- Public
- Created
- May 27, 2025
- Updated
- May 27, 2025
- SHA
- 238df115
Try a prompt
Model card
From the Hugging Face model README
Model Card for Model ID
class Config: base_model: str = "MattBou00/SmolLM-toxic-finetuned" reward_model: str = "s-nlp/roberta_toxicity_classifier" dataset_name: str = "skrishna/toy-toxicity-dataset" dataset_split: str = "test" prompt_words: int = 2 # use first N words of sample max_new_tokens: int = 16 sample_temperature: float = 1.0 top_p: float = 0.9 lr: float = 2e-6 # very low LR for stability epochs: int = 1 # bump for stronger learning entropy_coef: float = 0.005 save_dir: str = "./smollm_detox_rlhf" hub_repo: str = "MattBou00/SmolLM-toxic-detox-rlhf" # change if you like
- Task
- Text Generation
- Library
- transformers
- Type
- llama
- Created
- May 27, 2025
- Updated
- May 27, 2025