Instructions to use freewheelin/free-evo-qwen72b-v0.8-re with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use freewheelin/free-evo-qwen72b-v0.8-re with Transformers:

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

pipe = pipeline("text-generation", model="freewheelin/free-evo-qwen72b-v0.8-re")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("freewheelin/free-evo-qwen72b-v0.8-re")
model = AutoModelForCausalLM.from_pretrained("freewheelin/free-evo-qwen72b-v0.8-re")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use freewheelin/free-evo-qwen72b-v0.8-re with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "freewheelin/free-evo-qwen72b-v0.8-re"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "freewheelin/free-evo-qwen72b-v0.8-re",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/freewheelin/free-evo-qwen72b-v0.8-re

SGLang

How to use freewheelin/free-evo-qwen72b-v0.8-re 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 "freewheelin/free-evo-qwen72b-v0.8-re" \
    --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": "freewheelin/free-evo-qwen72b-v0.8-re",
		"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 "freewheelin/free-evo-qwen72b-v0.8-re" \
        --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": "freewheelin/free-evo-qwen72b-v0.8-re",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use freewheelin/free-evo-qwen72b-v0.8-re with Docker Model Runner:
```
docker model run hf.co/freewheelin/free-evo-qwen72b-v0.8-re
```

Model Card for free-evo-qwen72b-v0.8

Developed by : Freewheelin AI Technical Team

2024 4th May - avg. 81.28 Open Llm Leaderboard

Metric	Value
Avg.	81.28
ARC (25-Shot)	79.86
HellaSwag (10-Shot)	91.32
MMLU (5-Shot)	78.00
TruthfulQA (0-shot)	74.85
Winogrande (5-shot)	87.77
GSM8k (5-shot)	75.89

Method

We were inspired by this Sakana project

Process

You need two models with the same architecture.

Choose one model and fine-tune it to create a gap between the original model and the fine-tuned one. It doesn't matter whether the evaluation score is higher or lower.
Merge the two models.
Evaluate the merged model.
Fine-tune a specific evaluation part of the model if you need to increase the score for that part. (It's unlikely to work as you think, but you can try it.)
Merge the models again.
Evaluate again.
Keep going until the average evaluation score is higher than the original one.

That's it. Simple. You can create a framework to automate this process.

Base Architecture

QWEN2

Base Models

several QWEN2 based models

Downloads last month: 8,378

Safetensors

Model size

72B params

Tensor type

F16

Model tree for freewheelin/free-evo-qwen72b-v0.8-re

Merges

2 models

Space using freewheelin/free-evo-qwen72b-v0.8-re 1

Evaluation results

normalized accuracy on AI2 Reasoning Challenge (25-Shot)
test set Open LLM Leaderboard

79.860
normalized accuracy on HellaSwag (10-Shot)
validation set Open LLM Leaderboard

91.340
accuracy on MMLU (5-Shot)
test set Open LLM Leaderboard

78.000
mc2 on TruthfulQA (0-shot)
validation set Open LLM Leaderboard

74.850
accuracy on Winogrande (5-shot)
validation set Open LLM Leaderboard

87.770
accuracy on GSM8k (5-shot)
test set Open LLM Leaderboard

75.890