Calme 3.2 Instruct 78b

🚀 MaziyarPanahi/calme-3.2-instruct-78b

This model is an advanced iteration of Qwen/Qwen2.5-72B, fine - tuned for better performance in generic domains.

💡 Usage Tip

This is an experimental model, so it might not perform well for some prompts and may be sensitive to hyper parameters. I would appreciate any feedback to see if I can fix any issues in the next iteration. ❤️

🚀 Quick Start

This model is an advanced iteration of the powerful Qwen/Qwen2.5-72B, specifically fine - tuned to enhance its capabilities in generic domains. The Qwen2.5-72B base model was merged with itself to create a larger model. After that, the model was fine - tuned on a custom datasets.

✨ Features

• Advanced Iteration: Based on the powerful Qwen/Qwen2.5-72B, with self - merging and fine - tuning on custom datasets.
• Quantized Versions: Available in GGUF and EXL2 quantized formats for more efficient use.

📦 Installation

⚡ Quantized GGUF

Here are the GGUF models thanks to bartowski: calme-3.2-instruct-78b-GGUF

⚡ Quantized EXL2

Here is the EXL2 4.5 bits per weight (bpw) model thanks to DavidCatalano: DavidCatalano/calme-3.2-instruct-78b-exl2

DavidCatalano/calme-3.2-instruct-78b-exl2-4.5bpw.

💻 Usage Examples

Basic Usage

# Use a pipeline as a high - level helper

from transformers import pipeline

messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe = pipeline("text-generation", model="MaziyarPanahi/calme-3.2-instruct-78b")
pipe(messages)

Advanced Usage

# Load model directly

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("MaziyarPanahi/calme-3.2-instruct-78b")
model = AutoModelForCausalLM.from_pretrained("MaziyarPanahi/calme-3.2-instruct-78b")

📚 Documentation

🏆 Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Prompt Template

This model uses ChatML prompt template:

<|im_start|>system
{System}
<|im_end|>
<|im_start|>user
{User}
<|im_end|>
<|im_start|>assistant
{Assistant}

Ethical Considerations

As with any large language model, users should be aware of potential biases and limitations. We recommend implementing appropriate safeguards and human oversight when deploying this model in production environments.

📄 License

This model uses the qwen license. You can find the detailed license information here.