Delentia

delentia-lora-router-v0.4

README

License: apache-2.0

⚡ Quick Start: Load Adapter via PEFT

To execute this specialized classification adapter, load it on top of the base model:

python
from transformers import AutoModelForSequenceClassification, AutoTokenizer
from peft import PeftModel

base_model_name = "Delentia/delentia-slm-jitna-v0.4"
adapter_name = "Delentia/delentia-lora-router-v0.4"

# Load base model & tokenizer
model = AutoModelForSequenceClassification.from_pretrained(base_model_name, num_labels=4)
tokenizer = AutoTokenizer.from_pretrained(base_model_name)

# Load adapter
model = PeftModel.from_pretrained(model, adapter_name)

🌐 Delentia OS Ecosystem Model Roster (v0.4.x)

Delentia OS is organized into two primary deployment styles: Dynamic PEFT Adapters (1+4 Pillars) for sub-ms switching in unified VRAM, and Pre-Merged GGUF Models for direct plug-and-play local execution in Ollama / llama.cpp.

Table with columns: Component / Role, Deployment Type, Hugging Face Repository, Description, GGUF Support
Component / Role	Deployment Type	Hugging Face Repository	Description	GGUF Support
SLM Base Kernel	Base Foundation	Delentia/delentia-slm-jitna-v0.4	Core cognitive LLM (8B Parameters)	✅
The Router	PEFT LoRA Adapter	Delentia/delentia-lora-router-v0.4	Intention parser & node routing	❌ (PEFT only)
The Executor	PEFT LoRA Adapter	Delentia/delentia-lora-executor-v0.4	JSON tool payload generation	✅ (Merged GGUF below)
The Guardian	PEFT LoRA Adapter	Delentia/delentia-lora-guardian-v0.4	Zero-trust constitutional safety	✅ (Merged GGUF below)
The Scribe	PEFT LoRA Adapter	Delentia/delentia-lora-scribe-v0.4	Context compression/summarization	✅ (Merged GGUF below)
Pre-Merged Executor	Pre-Merged GGUF	Delentia/delentia-slm-jitna-executor-v0.4	Complete tool executor (plug-and-play)	✅
Pre-Merged Guardian	Pre-Merged GGUF	Delentia/delentia-slm-jitna-guardian-v0.4	Full safety guardrail model	✅
Pre-Merged Scribe	Pre-Merged GGUF	Delentia/delentia-slm-jitna-scribe-v0.4	Out-of-the-box context compressor	✅

Technical Specifications

Base Model: unsloth/Meta-Llama-3.1-8B-bnb-4bit
Fine-Tuning Method: Sequence Classification QLoRA (SEQ_CLS adapter config)
Target Modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj
Output Labels:
- 0: Router Base (Conversational / Standard Prompt)
- 1: Executor (Tool / JSON Execution)

Certified GPU Runs (v0.4 Performance)

Routing Classification Accuracy: 100.00% (Target Gate: $\geq 96.0%$ )
VRAM Swap Latency: 11.2 milliseconds (Target Gate: $\leq 12.0 ms$ )
Inference Speed: 20-50 milliseconds on consumer-grade local hardware.

🔒 Empirical Audit Ledger

The domain-specific empirical results below were generated and certified via system digital forensics:

Empirical Performance Graph

Auditor Notebook: colab_4_pillars_v043.ipynb (GitHub Source) |
Run ID: a5541e6e
Target Safetensors Hash: SHA256:0a1a9dec583c2259904b7504fd3483ea47563bce5257f5740d197ada173cb83a
Last Certified: 2026-07-13T06:26:20Z

Table with columns: Gate Category, Specific Metric, Target, Empirical Result, Status
Gate Category	Specific Metric	Target	Empirical Result	Status
Silicon Attestation	PCIe VRAM Swap Latency	< 12.0 ms	11.2000 ms	Certified (Cloud)
Cognitive Routing	Intent Classification Accuracy	>= 96.00%	100.00%	Certified
Economic Gate	API Cost Reduction Ratio	>= 90.00%

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Container

Run this model inference with full control and performance in your environment.

Learn more

Model Details

Model Provider

Delentia

Model Tree

Base

Delentia/delentia-slm-jitna-v0.4

Adapter

this model

Input Modalities

Text

Output Modalities