We define the text memory skill which we use for this example. We have chosen to use volatile memory, which uses the in-machine memory. Kernel.import_skill(sk.core_skills.TextMemorySkill()) Kernel.register_memory_store(memory_store=sk.memory.VolatileMemoryStore()) Sk_hf.HuggingFaceTextEmbedding("sentence-transformers/all-MiniLM-L6-v2"), "sentence-transformers/all-MiniLM-L6-v2", "gpt2", sk_hf.HuggingFaceTextCompletion("gpt2", task="text-generation") In this example we will use gp2 for text completion and sentence-transformers/all-MiniLM-L6-v2 for text embeddings. Next, we create a kernel instance and configure the hugging face services we want to use. !python -m pip install -r requirements.txt In the first two cells we install the relevant packages with a pip install and import the Semantic Kernel dependances. This video will give you a walk-through how to get started or dive right into the Python Sample here. For the remainder of this blog post we will be using the Hugging Face Sample with Skills as reference. How to use Semantic Kernel with Hugging Face? Using Hugging Face model services can provide great efficiencies as models are pre-trained, easy to swap out and cost-effective with many free models available. Models are pre-trained on large datasets and can be used to quickly perform a variety of tasks, such as sentiment analysis, text classification, and text summarization. Hugging Face is a leading provider of open-source models. This integration allows you to use the vast number of models at your fingertips with the latest advancements in Semantic Kernel’s orchestration, skills, planner and contextual memory support. With this integration, you can leverage the power of Semantic Kernel combined with accessibility of over 190,000+ models from Hugging Face. We are thrilled to announce the integration of Semantic Kernel with Hugging Face models!
0 kommentar(er)
