Panth-L3-Blackroot-Nephra-8B-MNN

Pre-converted Panth-L3-Blackroot-Nephra-MK.VI-8B in MNN format for on-device inference with TokForge.

Original model by mergekit-community — converted to MNN Q4 for mobile deployment.

Model Details

Architecture	Llama 3 (standard attention, 32 layers, GQA 32Q/8KV)
Parameters	8B (4-bit quantized)
Format	MNN (Alibaba Mobile Neural Network)
Quantization	W4A16 (4-bit weights, block size 128)
Vocab	128,256 tokens
Source	mergekit-community/Panth-L3-Blackroot-Nephra-MK.VI-8B

Description

Panth-L3 Blackroot Nephra MK.VI — a Llama 3 8B merge by mergekit-community. Blends multiple fine-tunes for strong creative writing and roleplay capabilities.

Files

File	Description
llm.mnn	Model computation graph
llm.mnn.weight	Quantized weight data (Q4, block=128)
llm_config.json	Model config with Jinja chat template
tokenizer.txt	Tokenizer vocabulary
config.json	MNN runtime config

Usage with TokForge

This model is optimized for TokForge — a free Android app for private, on-device LLM inference.

1. Download TokForge from the Play Store
2. Open the app → Models → Download this model
3. Start chatting — runs 100% locally, no internet required

Recommended Settings

Setting	Value
Backend	OpenCL (Qualcomm) / Vulkan (MediaTek) / CPU (fallback)
Precision	Low
Threads	4
Thinking	Off (or On for thinking-capable models)

Performance

Actual speed varies by device, thermal state, and generation length. Typical ranges for this model size:

Device	SoC	Backend	tok/s
RedMagic 11 Pro	SM8850	OpenCL	14.7 tok/s

Attribution

This is an MNN conversion of Panth-L3-Blackroot-Nephra-MK.VI-8B by mergekit-community. All credit for the model architecture, training, and fine-tuning goes to the original author(s). This conversion only changes the runtime format for mobile deployment.

Limitations

• Intended for TokForge / MNN on-device inference on Android
• This is a runtime bundle, not a standard Transformers training checkpoint
• Quantization (Q4) may slightly reduce quality compared to the full-precision original
• Abliterated/uncensored models have had safety filters removed — use responsibly

Community

• Website: tokforge.ai
• Discord: Join our Discord
• GitHub: TokForge on GitHub

Export Details

Converted using MNN's llmexport pipeline:

python llmexport.py --path mergekit-community/Panth-L3-Blackroot-Nephra-MK.VI-8B --export mnn --quant_bit 4 --quant_block 128