| license: apache-2.0 | |
| tags: | |
| - network-engineering | |
| - cisco | |
| - network-automation | |
| - fine-tuned | |
| - security | |
| language: | |
| - en | |
| base_model: mistralai/Mistral-7B-Instruct-v0.2 | |
| pipeline_tag: text-generation | |
| # Security - Fine-Tuned Network Agent | |
| Fine-tuned from `mistralai/Mistral-7B-Instruct-v0.2` for autonomous network operations. | |
| ## Model Details | |
| - **Agent Type**: security | |
| - **Base Model**: mistralai/Mistral-7B-Instruct-v0.2 | |
| - **Method**: LoRA fine-tuning (merged into base weights) | |
| - **Format**: safetensors | |
| - **Use Case**: Network incident analysis and troubleshooting | |
| ## Usage | |
| ```python | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| model = AutoModelForCausalLM.from_pretrained("eduard76/security-Mistral-7B-Instruct-v0.2") | |
| tokenizer = AutoTokenizer.from_pretrained("eduard76/security-Mistral-7B-Instruct-v0.2") | |
| prompt = "Analyze OSPF neighbor flapping issue" | |
| inputs = tokenizer(prompt, return_tensors="pt") | |
| outputs = model.generate(**inputs, max_new_tokens=512) | |
| print(tokenizer.decode(outputs[0])) | |
| ``` | |
| ## Training | |
| - **Method**: LoRA (rank 16-64) | |
| - **Dataset**: Network operations scenarios | |
| - **Target**: Cisco network operations | |
| ## License | |
| Apache 2.0 | |