Instructions to use bobofrut/ladybird-base-7B-v8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bobofrut/ladybird-base-7B-v8 with Transformers:

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

pipe = pipeline("text-generation", model="bobofrut/ladybird-base-7B-v8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("bobofrut/ladybird-base-7B-v8")
model = AutoModelForCausalLM.from_pretrained("bobofrut/ladybird-base-7B-v8")
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 Settings

vLLM

How to use bobofrut/ladybird-base-7B-v8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bobofrut/ladybird-base-7B-v8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bobofrut/ladybird-base-7B-v8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/bobofrut/ladybird-base-7B-v8

SGLang

How to use bobofrut/ladybird-base-7B-v8 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 "bobofrut/ladybird-base-7B-v8" \
    --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": "bobofrut/ladybird-base-7B-v8",
		"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 "bobofrut/ladybird-base-7B-v8" \
        --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": "bobofrut/ladybird-base-7B-v8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use bobofrut/ladybird-base-7B-v8 with Docker Model Runner:
```
docker model run hf.co/bobofrut/ladybird-base-7B-v8
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Ladybird-base-7B-v8

Welcome to the repository of Ladybird-base-7B-v8, a cutting-edge Large Language Model (LLM) developed as a result of extensive research and learning in the field of Artificial Intelligence (AI), particularly focusing on LLMs. This model represents a significant milestone in my journey to understand and contribute to the advancement of AI technologies.

About the Creator

As an avid learner and researcher of AI, I embarked on the journey to not only understand but also to contribute to the field of Large Language Models. Building and fine-tuning my own models allowed me to deeply engage with the intricacies of AI, culminating in the development of the Ladybird-base-7B-v8. This project is a testament to my dedication to learning and my passion for pushing the boundaries of what AI models can achieve.

Model Overview

Ladybird-base-7B-v8 is based on the Mistral architecture, which is known for its efficiency and effectiveness in handling complex language understanding and generation tasks. The model incorporates several innovative architecture choices to enhance its performance:

Grouped-Query Attention: Optimizes attention mechanisms by grouping queries, reducing computational complexity while maintaining model quality.
Sliding-Window Attention: Improves the model's ability to handle long-range dependencies by focusing on relevant segments of input, enhancing understanding and coherence.
Byte-fallback BPE Tokenizer: Offers robust tokenization by combining the effectiveness of Byte-Pair Encoding (BPE) with a fallback mechanism for out-of-vocabulary bytes, ensuring comprehensive language coverage.

Instruction Format

To fully leverage the capabilities of Ladybird-base-7B-v8, especially its instruction fine-tuning feature, users are advised to follow ChatML format. This format ensures that prompts are effectively processed, resulting in accurate and context-aware responses from the model. Here's how to construct your prompts:

msg = [
    {"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
    {"role": "assistant", "content": "You are a friendly chatbot who always responds in the style of a pirate"},
]

prompt = pipe.tokenizer.apply_chat_template(msg, tokenize=False, add_generation_prompt=True)

Eval results

Tasks	Version	Filter	n-shot	Metric	Value		Stderr
winogrande	1	none	None	acc	0.8272	±	0.0106
truthfulqa_mc2	2	none	0	acc	0.7736	±	0.0139
truthfulqa_mc1	2	none	0	acc	0.6242	±	0.0170
stem	N/A	none	None	acc	0.5109	±	0.0085
- abstract_algebra	0	none	None	acc	0.2900	±	0.0456
- anatomy	0	none	None	acc	0.5852	±	0.0426
- astronomy	0	none	None	acc	0.6908	±	0.0376
- college_biology	0	none	None	acc	0.6875	±	0.0388
- college_chemistry	0	none	None	acc	0.4000	±	0.0492
- college_computer_science	0	none	None	acc	0.5300	±	0.0502
- college_mathematics	0	none	None	acc	0.2600	±	0.0441
- college_physics	0	none	None	acc	0.4314	±	0.0493
- computer_security	0	none	None	acc	0.7100	±	0.0456
- conceptual_physics	0	none	None	acc	0.5702	±	0.0324
- electrical_engineering	0	none	None	acc	0.5586	±	0.0414
- elementary_mathematics	0	none	None	acc	0.4259	±	0.0255
- high_school_biology	0	none	None	acc	0.7710	±	0.0239
- high_school_chemistry	0	none	None	acc	0.4483	±	0.0350
- high_school_computer_science	0	none	None	acc	0.7000	±	0.0461
- high_school_mathematics	0	none	None	acc	0.3259	±	0.0286
- high_school_physics	0	none	None	acc	0.3179	±	0.0380
- high_school_statistics	0	none	None	acc	0.4491	±	0.0339
- machine_learning	0	none	None	acc	0.5000	±	0.0475
hellaswag	1	none	None	acc	0.7010	±	0.0046
		none	None	acc_norm	0.8763	±	0.0033
gsm8k	3	strict-match	5	exact_match	0.7650	±	0.0117
		flexible-extract	5	exact_match	0.7695	±	0.0116
arc_challenge	1	none	None	acc	0.6749	±	0.0137
		none	None	acc_norm	0.6800	±	0.0136

Contact

jackiewicz.raf@gmail.com

Downloads last month: 81

Safetensors

Model size

7B params

Tensor type

BF16

Model tree for bobofrut/ladybird-base-7B-v8

Adapters

1 model

Merges

1 model

Quantizations

4 models