Instructions to use MetaCore-LLM/MetaCore-1-Test-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MetaCore-LLM/MetaCore-1-Test-Instruct with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="MetaCore-LLM/MetaCore-1-Test-Instruct")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MetaCore-LLM/MetaCore-1-Test-Instruct")
model = AutoModelForCausalLM.from_pretrained("MetaCore-LLM/MetaCore-1-Test-Instruct", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use MetaCore-LLM/MetaCore-1-Test-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MetaCore-LLM/MetaCore-1-Test-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MetaCore-LLM/MetaCore-1-Test-Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/MetaCore-LLM/MetaCore-1-Test-Instruct

SGLang

How to use MetaCore-LLM/MetaCore-1-Test-Instruct 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 "MetaCore-LLM/MetaCore-1-Test-Instruct" \
    --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": "MetaCore-LLM/MetaCore-1-Test-Instruct",
		"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 "MetaCore-LLM/MetaCore-1-Test-Instruct" \
        --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": "MetaCore-LLM/MetaCore-1-Test-Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use MetaCore-LLM/MetaCore-1-Test-Instruct with Docker Model Runner:
```
docker model run hf.co/MetaCore-LLM/MetaCore-1-Test-Instruct
```

MetaCore-1-Test-Instruct

File size: 700 Bytes

---
language:
- ru
license: mit
library_name: transformers
pipeline_tag: text-generation
tags:
- russian
- llama
- causal-lm
- instruction-tuning
- sft
- tiny
base_model: MetaCore-LLM/MetaCore-1-Test-CPT
datasets:
- d0rj/ru-instruct
model-index:
- name: MetaCore-1-Test-Instruct
  results: []
---

# MetaCore-1-Test-Instruct

**MetaCore-1-Test-Instruct** is a lightweight Russian-language instruction-tuned model, built on top of the **MetaCore-1-Test-CPT** foundation model. It is intended for experimentation, prototyping, and educational purposes.

- **Developer:** MetaCore-LLM
- **Architecture:** Custom LLaMA-style (tiny config)
- **Language:** Russian
- **Parameter count:** ~16.2 million

---