ipex-llm
ipex-llm copied to clipboard
intel-analytics
Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, Baichuan, Mixtral, Gemma, Phi, etc.) on Intel CPU and GPU (e.g., local PC with iGPU, discrete GPU such as Arc, Flex and Ma...
## Description forecaster user guide how to save and load forecaster
Issue 1: `to_tf_dataset` has no description about `shuffle`. https://bigdl.readthedocs.io/en/latest/doc/PythonAPI/Chronos/tsdataset.html#bigdl.chronos.data.tsdataset.TSDataset.to_tf_dataset Issue 2: Currently we support six built-in datasets, but lack the details of 'uci_electricity_wideare'. https://bigdl.readthedocs.io/en/latest/doc/Chronos/Overview/data_processing_feature_engineering.html#built-in-dataset
When running `source bigdl-nano-init` in bash, it usually works well, but when running it in other shells like zsh, the script has trouble setting the preload libs. I think it's...
## Description The method argument is set to a `functools.partial` when training a yolox model with multiple processes. ## Traceback ``` Traceback (most recent call last): File "/home/mingzhi/anaconda3/envs/nano_yolox/lib/python3.7/site-packages/torch/multiprocessing/spawn.py", line 69,...
Hi, I'm running into this error whenever I set the `class_weight` parameter. The demo script below replicates the error. If the `class_weight` parameter is removed then the code completes without...
## Description Add FL API doc, continuing updating. add API doc using the style of [Orca estimator](https://github.com/intel-analytics/BigDL/blob/main/python/orca/src/bigdl/orca/tfpark/estimator.py) for all the API level methods, including following @jason-dai please directly edit the...
For third-party dataloader implementations, we can use a `collate_fn` wrapper to implement custom data processing like elements alignment: ``` def reduceWrapper(func): def reduceIterElement(batch): batch=func(batch) assert len(batch)==5, "Should yield inputs, labels,...
We need such a how to guide to help users to understand how to predict future data by using a trained forecaster Critical APIs includes `forecaster.predict` `tsdataset.unscale_numpy()` (and why it...
## Description * set jemalloc as the default allocator
Metadata
Owner
Metadata
Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, Baichuan, Mixtral, Gemma, Phi, etc.) on Intel CPU and GPU (e.g., local PC with iGPU, discrete GPU such as Arc, Flex and Ma...