Please use in accordance with Llama-2's license terms.

For better results, the Prompt should be as following:

A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: 

Introduction

WizardLM-70B V1.0 is a large language model, trained from Llama-2 70b. on AI-evolved instructions using the Evol+ approach. This model is designed to follow complex instructions and generate coherent and fluent text in response to various inputs. The model is trained on a large volume of text data and is capable of achieving high performance on various natural language processing tasks. 

WizardLM-70B Model

WizardLM-70B V1.2 is a transformer-based language model with 70 billion parameters. It is fine-tuned on AI-evolved instructions using the Evol+ approach. The model is pre-trained on a large corpus of text data and fine-tuned on the Llama-2 dataset to generate high-quality responses to complex instructions. The model supports a 4k context window and is licensed under the same terms as Llama-2. 

Use Cases

Question Answering

Prompt:
What is the capital of France?
Output:
The capital of France is Paris.

Language Translation

Prompt:
Translate "Bonjour, comment ça va?" to English.
Output:
Hello, how are you?

Text Generation

Prompt:
Write a short story about a boy who discovers a magical world.
Output:
Once upon a time, there was a boy named Jack who stumbled upon a magical world hidden behind a secret door in his backyard...

Dataset Information

WizardLM-70B is fine-tuned on the Llama-2 dataset, which is a collection of AI-evolved instructions using the Evol+ approach. The dataset is designed to test the ability of language models to follow complex instructions and generate high-quality responses. The dataset contains a variety of instructions, ranging from simple to complex, and covers a wide range of natural language processing tasks. 

Manually creating instruction data is really time-consuming and humans are lazy and are not consistent. Evol-Instruct is a methodology to develop complex instructions (evolving instructions from less complex to more) using LLMs. WizardLM was trained using a dataset generated using Evol-Instruct.

Evaluation

WizardLM-70B V1.2 is evaluated using both automatic and human evaluation metrics. The model is evaluated on various natural language processing tasks, including question answering, language translation, and text generation. The model achieves impressive results, with a score of 7.78 on MT-Bench, 92.91% on Alpaca Eval, 77.6% on GSM8k and 50.6 on HumanEval.

 Limitations

Despite its high performance on various natural language processing tasks, WizardLM-70B V1.2 has some limitations. The model may not perform well on tasks that require domain-specific knowledge or on tasks that

Disclaimer

Please be advised that this model utilizes wrapped Artificial Intelligence (AI) provided by TogetherAI (the "Vendor"). These AI models may collect, process, and store data as part of their operations. By using our website and accessing these AI models, you hereby consent to the data practices of the Vendor. We do not have control over the data collection, processing, and storage practices of the Vendor. Therefore, we cannot be held responsible or liable for any data handling practices, data loss, or breaches that may occur. It is your responsibility to review the privacy policies and terms of service of the Vendor to understand their data practices. You can access the Vendor's privacy policy and terms of service at https://www.togetherai.com/legal/privacy-policy.

We disclaim all liability with respect to the actions or omissions of the Vendor, and we encourage you to exercise caution and to ensure that you are comfortable with these practices before utilizing the AI models hosted on our site.