Instructions to use 17Lab/qwen14b-dpo-rank256-s42 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 17Lab/qwen14b-dpo-rank256-s42 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("17Lab/qwen14b-strict-json-full-sft-s42")
model = PeftModel.from_pretrained(base_model, "17Lab/qwen14b-dpo-rank256-s42")

Transformers

How to use 17Lab/qwen14b-dpo-rank256-s42 with Transformers:

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

pipe = pipeline("text-generation", model="17Lab/qwen14b-dpo-rank256-s42")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("17Lab/qwen14b-dpo-rank256-s42", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use 17Lab/qwen14b-dpo-rank256-s42 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "17Lab/qwen14b-dpo-rank256-s42"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "17Lab/qwen14b-dpo-rank256-s42",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/17Lab/qwen14b-dpo-rank256-s42

SGLang

How to use 17Lab/qwen14b-dpo-rank256-s42 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 "17Lab/qwen14b-dpo-rank256-s42" \
    --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": "17Lab/qwen14b-dpo-rank256-s42",
		"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 "17Lab/qwen14b-dpo-rank256-s42" \
        --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": "17Lab/qwen14b-dpo-rank256-s42",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use 17Lab/qwen14b-dpo-rank256-s42 with Docker Model Runner:
```
docker model run hf.co/17Lab/qwen14b-dpo-rank256-s42
```

qwen14b_dpo_from_sft_rank_s42

This model is a fine-tuned version of Qwen/Qwen2.5-14B on the assimilation_dpo_v1 dataset. It achieves the following results on the evaluation set:

Loss: 0.0272
Rewards/chosen: -0.1747
Rewards/rejected: -6.2640
Rewards/accuracies: 1.0
Rewards/margins: 6.0893
Logps/chosen: -5.2166
Logps/rejected: -80.8057
Logits/chosen: -1.5318
Logits/rejected: -1.5291

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-06
train_batch_size: 2
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 2
gradient_accumulation_steps: 16
total_train_batch_size: 64
total_eval_batch_size: 2
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/chosen	Logps/rejected	Logits/chosen	Logits/rejected
0.0329	1.0	25	0.0272	-0.1747	-6.2640	1.0	6.0893	-5.2166	-80.8057	-1.5318	-1.5291

Framework versions

PEFT 0.18.1
Transformers 5.6.0
Pytorch 2.7.0+cu128
Datasets 4.0.0
Tokenizers 0.22.2

Downloads last month: 2

Model tree for 17Lab/qwen14b-dpo-rank256-s42

Base model

Qwen/Qwen2.5-14B

Adapter

(29)

this model