Instructions to use sayeed105236/cuteadmoa-lora-adapter with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use sayeed105236/cuteadmoa-lora-adapter with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-14B-Instruct")
model = PeftModel.from_pretrained(base_model, "sayeed105236/cuteadmoa-lora-adapter")

Transformers

How to use sayeed105236/cuteadmoa-lora-adapter with Transformers:

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

pipe = pipeline("text-generation", model="sayeed105236/cuteadmoa-lora-adapter")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("sayeed105236/cuteadmoa-lora-adapter", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use sayeed105236/cuteadmoa-lora-adapter with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "sayeed105236/cuteadmoa-lora-adapter"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sayeed105236/cuteadmoa-lora-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/sayeed105236/cuteadmoa-lora-adapter

SGLang

How to use sayeed105236/cuteadmoa-lora-adapter 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 "sayeed105236/cuteadmoa-lora-adapter" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sayeed105236/cuteadmoa-lora-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "sayeed105236/cuteadmoa-lora-adapter" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "sayeed105236/cuteadmoa-lora-adapter",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use sayeed105236/cuteadmoa-lora-adapter with Docker Model Runner:
```
docker model run hf.co/sayeed105236/cuteadmoa-lora-adapter
```

CuttyMOA-1.0 LoRA Adapter

This is the LoRA adapter used to train CuttyMOA-1.0, the first self-trained model from the CUTEADMOA platform.

Training Details

Base model: Qwen/Qwen2.5-14B-Instruct
Method: QLoRA 4-bit (nf4, double quantization)
LoRA config: r=128, α=256, dropout=0.05
Target modules: q_proj, k_proj, v_proj, o_proj, gate_proj, up_proj, down_proj (7 modules)
Training steps: ~20,000 (RunPod A40 48GB)
Dataset: sayeed105236/cuteadmoa-204-datasets (4.5M pairs, 24 categories)

Two-Phase Training

This adapter represents Phase 2 (20K steps on RunPod). Phase 1 (4K steps on HuggingFace) produced an earlier version of this adapter at commit a26500121. Both phases were TIES-merged into the final CuttyMOA-1.0 model.

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-14B-Instruct", torch_dtype="float16")
model = PeftModel.from_pretrained(base, "sayeed105236/cuteadmoa-lora-adapter")
model = model.merge_and_unload()

Downloads last month: 64

Model tree for sayeed105236/cuteadmoa-lora-adapter

Base model

Qwen/Qwen2.5-14B

Finetuned

Qwen/Qwen2.5-14B-Instruct

Adapter

(392)

this model