search

intel-analytics / ipex-llm

Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, Baichuan, Mixtral, Gemma, Phi, MiniCPM, etc.) on Intel CPU and GPU (e.g., local PC with iGPU, discrete GPU such as Arc, Flex and Max); seamlessly integrate with llama.cpp, Ollama, HuggingFace, LangChain, LlamaIndex, GraphRAG, DeepSpeed, vLLM, FastChat, Axolotl, etc.
Apache License 2.0
6.45k stars 1.24k forks source link

Text Classification Example needs prediction part #1522

Open hhbyyh opened 7 years ago

yiheng commented 7 years ago

@hhbyyh Does this meet your requirement. Basically, it trains a text classifier model and uses it in DataFrame query.

https://github.com/intel-analytics/BigDL/blob/master/spark/dl/src/main/scala/com/intel/analytics/bigdl/example/udfpredictor/DataframePredictor.scala

hhbyyh commented 7 years ago

Hi @yiheng The complain actually comes from the customer Gigaspace. I guess when they could not find the prediction part for the Text classification example. They tried to implement something by themselves and met some problems.

IMO, we may either link the udfPredictor to the text classification, or implement the rdd-based prediction part for text classification.

yiheng commented 7 years ago

I see. There's an udf example in the BigDL-Tutorial. See this PR: https://github.com/intel-analytics/BigDL-Tutorials/pull/20. But I feel it's too complex. Does it meet your requirement?

@yangw1234 Can we make the udf example simpler? like on scala notebook?

yangw1234 commented 7 years ago

@yiheng I'll try and do that.

yiheng commented 7 years ago

Great, we're trying to add scala notebook(backend is toree) to BigDL-Tutorial. It will be much more friendly for user to learn how to use BigDL through notebook. One thing, I'm not sure how to handle java package dependency in toree based scala notebook. Is there something like pip in JVM?

yiheng commented 7 years ago

Or we can use mvn repository as java class path