Instructions to use realPCH/kosolra_SFT_DPO_v0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use realPCH/kosolra_SFT_DPO_v0 with Transformers:

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

pipe = pipeline("text-generation", model="realPCH/kosolra_SFT_DPO_v0")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("realPCH/kosolra_SFT_DPO_v0")
model = AutoModelForCausalLM.from_pretrained("realPCH/kosolra_SFT_DPO_v0")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use realPCH/kosolra_SFT_DPO_v0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "realPCH/kosolra_SFT_DPO_v0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "realPCH/kosolra_SFT_DPO_v0",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/realPCH/kosolra_SFT_DPO_v0

SGLang

How to use realPCH/kosolra_SFT_DPO_v0 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 "realPCH/kosolra_SFT_DPO_v0" \
    --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": "realPCH/kosolra_SFT_DPO_v0",
		"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 "realPCH/kosolra_SFT_DPO_v0" \
        --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": "realPCH/kosolra_SFT_DPO_v0",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use realPCH/kosolra_SFT_DPO_v0 with Docker Model Runner:
```
docker model run hf.co/realPCH/kosolra_SFT_DPO_v0
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Developed by chPark

Training Strategy

We fine-tuned this model based on yanolja/KoSOLAR-10.7B-v0.1 We applied a DPO to realPCH/kosolra-kullm

Run the model

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "realPCH/kosolra_SFT_DPO_v0"
tokenizer = AutoTokenizer.from_pretrained(model_id)

model = AutoModelForCausalLM.from_pretrained(model_id)

text = "[INST] Put instruction here. [/INST]"
inputs = tokenizer(text, return_tensors="pt")

outputs = model.generate(**inputs, max_new_tokens=20)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Downloads last month: 5

Safetensors

Model size

11B params

Tensor type

F32

Model tree for realPCH/kosolra_SFT_DPO_v0

Quantizations

1 model

realPCH
/

kosolra_SFT_DPO_v0

Developed by chPark

Training Strategy

Run the model

Model tree for realPCH/kosolra_SFT_DPO_v0

Datasets used to train realPCH/kosolra_SFT_DPO_v0