Instructions to use kevin009/minirewrite with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kevin009/minirewrite with Transformers:

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

pipe = pipeline("text-generation", model="kevin009/minirewrite")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("kevin009/minirewrite")
model = AutoModelForCausalLM.from_pretrained("kevin009/minirewrite")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use kevin009/minirewrite with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kevin009/minirewrite"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kevin009/minirewrite",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kevin009/minirewrite

SGLang

How to use kevin009/minirewrite 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 "kevin009/minirewrite" \
    --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": "kevin009/minirewrite",
		"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 "kevin009/minirewrite" \
        --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": "kevin009/minirewrite",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use kevin009/minirewrite with Docker Model Runner:
```
docker model run hf.co/kevin009/minirewrite
```

kevin009 commited on Jul 7, 2024

Commit

622b387

verified ·

1 Parent(s): ae59e8b

Update README.md

Browse files

Files changed (1) hide show

README.md +28 -8

README.md CHANGED Viewed

@@ -1,22 +1,42 @@
 ---
-base_model: unsloth/mistral-7b-instruct-v0.2-bnb-4bit
 language:
 - en
 license: apache-2.0
 tags:
 - text-generation-inference
 - transformers
-- unsloth
 - mistral
 - trl
 ---
-# Uploaded  model
-- **Developed by:** kevin009
-- **License:** apache-2.0
-- **Finetuned from model :** unsloth/mistral-7b-instruct-v0.2-bnb-4bit
-This mistral model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)

 ---
 language:
 - en
 license: apache-2.0
 tags:
 - text-generation-inference
 - transformers
 - mistral
 - trl
 ---
+# Model Card: Minimalist Assistant
+## Model Details
+- **Architecture**: 32k tokens, 32 layers
+- **Quantization**: 4-bit
+- **Base Model**: Mistral Instruct
+- **Tokenizer**: Custom (based on Mistral Instruct)
+## Intended Use
+- As Editor Assistant for revision and paraphrasing
+## Training Data
+- **Initial Training**: 14,000 conversations in minimalist style to ensure concise output
+- **Further Training**: 8,000 revision conversations to enhance rewriting and paraphrasing capabilities
+## Performance and Limitations
+- **Strengths**:
+  - Optimized for generating concise content
+  - Specialized in rewriting and paraphrasing tasks
+- **Limitations**:
+  - May produce shorter outputs compared to standard models
+  - Potential biases from training data should be considered
+## Ethical Considerations
+- Designed for daily use, potential biases from training data should be considered
+- Users should be aware of the model's focus on brevity and rewriting
+## Additional Information
+- Fine-tuned to address limitations in writing tasks observed in other models
+- Personalized for everyday use cases
+- Motivation for development was to create a model better suited for writing tasks, as existing models were found lacking in this area