Text Generation
Transformers
Safetensors
English
Bengali
mistral
Merge
mergekit
lazymergekit
sillytavern
roleplay
rp
math
coding
12b
text-generation-inference
Instructions to use kainatq/RP-king-12b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use kainatq/RP-king-12b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="kainatq/RP-king-12b")# Load model directly from transformers import AutoTokenizer, AutoModelForMultimodalLM tokenizer = AutoTokenizer.from_pretrained("kainatq/RP-king-12b") model = AutoModelForMultimodalLM.from_pretrained("kainatq/RP-king-12b") - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use kainatq/RP-king-12b with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "kainatq/RP-king-12b" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "kainatq/RP-king-12b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/kainatq/RP-king-12b
- SGLang
How to use kainatq/RP-king-12b with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "kainatq/RP-king-12b" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "kainatq/RP-king-12b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "kainatq/RP-king-12b" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "kainatq/RP-king-12b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use kainatq/RP-king-12b with Docker Model Runner:
docker model run hf.co/kainatq/RP-king-12b
File size: 1,713 Bytes
3e3c91d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 | ---
library_name: transformers
license: apache-2.0
tags:
- merge
- mergekit
- lazymergekit
base_model:
- MarinaraSpaghetti/NemoMix-Unleashed-12B
- inflatebot/MN-12B-Mag-Mell-R1
- elinas/Chronos-Gold-12B-1.0
- Retreatcost/KansenSakura-Eclipse-RP-12b
- redrix/patricide-12B-Unslop-Mell-v2
---
# RP-king-12b
RP-king-12b is a merge of the following models using [mergekit](https://github.com/cg123/mergekit):
* [MarinaraSpaghetti/NemoMix-Unleashed-12B](https://huggingface.co/MarinaraSpaghetti/NemoMix-Unleashed-12B)
* [inflatebot/MN-12B-Mag-Mell-R1](https://huggingface.co/inflatebot/MN-12B-Mag-Mell-R1)
* [elinas/Chronos-Gold-12B-1.0](https://huggingface.co/elinas/Chronos-Gold-12B-1.0)
* [Retreatcost/KansenSakura-Eclipse-RP-12b](https://huggingface.co/Retreatcost/KansenSakura-Eclipse-RP-12b)
* [redrix/patricide-12B-Unslop-Mell-v2](https://huggingface.co/redrix/patricide-12B-Unslop-Mell-v2)
## 🧩 Configuration
```yamlmodels:
- model: MarinaraSpaghetti/NemoMix-Unleashed-12B # Base model for creative foundation
parameters:
density: 0.65
weight: 1.0
- model: inflatebot/MN-12B-Mag-Mell-R1 # For worldbuilding and prose
parameters:
density: 0.8
weight: 1.2
- model: elinas/Chronos-Gold-12B-1.0 # For roleplay and prompt adherence
parameters:
density: 0.9
weight: 1.1
- model: Retreatcost/KansenSakura-Eclipse-RP-12b
parameters:
density: 0.5
weight: 0.7
- model: redrix/patricide-12B-Unslop-Mell-v2 # For anti-GPTism and balance
parameters:
density: 0.6
weight: 0.9
merge_method: dare_ties
base_model: MarinaraSpaghetti/NemoMix-Unleashed-12B
parameters:
normalize: true
dtype: bfloat16
tokenizer_source: base``` |