Arxiv:2412.07724 Arxiv:2502.03397 Arxiv:2508.09937 Autotrain compatible Base model:finetune:ibm-granit... Base model:ibm-granite/granite... Conversational En Endpoints compatible Granite Lora Region:us Safetensors
Model Card on HF 🤗: https://huggingface.co/ibm-granite/granite-guardian-3.2-5b-lora-harm-correction
Granite Guardian 3.2 5B Lora Harm Correction Benchmarks
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
What is the LLM Explorer Rank (Score)
Granite Guardian 3.2 5B Lora Harm Correction Parameters and Internals
| LLM Name | Granite Guardian 3.2 5B Lora Harm Correction |
| Repository 🤗 | https://huggingface.co/ibm-granite/granite-guardian-3.2-5b-lora-harm-correction |
| Base Model(s) | |
| Model Size | 5b |
| Required VRAM | 0.1 GB |
| Updated | 2025-09-10 |
| Maintainer | ibm-granite |
| Model Files | |
| Supported Languages | en |
| Model Architecture | AutoModelForCausalLM |
| License | apache-2.0 |
| Model Max Length | 8192 |
| Is Biased | none |
| Tokenizer Class | GPT2Tokenizer |
| Padding Token | <|end_of_text|> |
| Vocabulary Size | 49152 |
| PEFT Type | LORA |
| LoRA Model | Yes |
| PEFT Target Modules | out_proj|q_proj|fc_out|v_proj|wte|k_proj|fc_in |
| LoRA Alpha | 32 |
| LoRA Dropout | 0.1 |
| R Param | 32 |
| Errors | replace |
Rank the Granite Guardian 3.2 5B Lora Harm Correction Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 51262 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Release v20241124