Zyrabit-Hub/zyra-multitask-master
Text Generation • Updated
None defined yet.
Enterprise Local Inference Engine & 3B SLM Suite
| Model ID | Domain / Function | Format | Size | Target Hardware |
|---|---|---|---|---|
| zyra-agent-hunter | B2B Signal Extraction & ICP Discovery | Q5_K_M | 2.1 GB | Tenstorrent Blackhole / CPU / GPU |
| zyra-agent-sentinel | In-RAM PII Sanitization & GDPR Audit | Q5_K_M | 2.1 GB | Tenstorrent Blackhole / CPU / GPU |
| zyra-agent-closer | Financial ROI & Token-Cost Calculation | Q5_K_M | 2.1 GB | Tenstorrent Blackhole / CPU / GPU |
| zyra-agent-strategist | Infrastructure Topology & NPU Mapping | Q5_K_M | 2.1 GB | Tenstorrent Blackhole / CPU / GPU |
| zyra-multitask-master | Unified Multi-Agent Orchestrator Node | Q5_K_M | 2.1 GB | Tenstorrent Blackhole / CPU / GPU |
The optimal and secure way to deploy Zyrabit SLMs in enterprise environments is via the official Zyrabit SLM Docker runtime.
docker pull zyrabitcore/zyrabit-slm:2.4.1
docker run -d \
--name zyrabit-api \
-p 8088:8080 \
-e INFERENCE_PROVIDER=ollama \
-e SLM_URL=http://host.docker.internal:11434 \
-e MODEL_NAME=zyra-agent-hunter \
-e ZYRABIT_API_KEY_WEB=your-secure-token \
zyrabitcore/zyrabit-slm:2.4.1
Full Sovereign Stack Execution (with Tenstorrent NPU support):
git clone https://github.com/Zyrabit-tech/zyrabit-SLM.git
cd zyrabit-SLM
docker compose --profile db --profile tenstorrent up -d
To prevent token looping, configure the ChatML template and stop tokens in your Ollama Modelfile:
FROM ./zyra-agent-hunter-Q5_K_M.gguf
TEMPLATE """<|im_start|>system
{{ .System }}<|im_end|>
<|im_start|>user
{{ .Prompt }}<|im_end|>
<|im_start|>assistant
{{ .Response }}<|im_end|>"""
SYSTEM """You are Zyra Agent Hunter, a specialized B2B outreach SLM by Zyrabit-IA."""
PARAMETER stop "<|im_start|>"
PARAMETER stop "<|im_end|>"
PARAMETER temperature 0.2
PARAMETER repeat_penalty 1.15
+---------------------------------------------------------------------------------+
| TRACEABILITY & HARDWARE BENCHMARK MATRIX |
+---------------------------------------------------------------------------------+
| * Architecture Base | Qwen2.5 (3B Parameters) |
| * Fine-Tuning Hardware | Tenstorrent Blackhole NPU (p150) / Arch.BLACKHOLE |
| * Dataset Verification | 2,000 Sanitized Expert Pairs (v1.0.0-sovereign) |
| * JSON Schema Compliance | 100.0% Structural Output Accuracy |
| * Air-Gap Network Egress | 0 Bytes Outbound (Isolated Execution) |
+---------------------------------------------------------------------------------+