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...
Summary: To merge two SparkXShards like `pandas.DataFrame.merge()`, add the function `merge()` to SparkXShards in the file `bigdl.orca.data.shard.py`. Also, a NCF pytorch example `train_xshards_add_features.py` is added to demonstrate how to use...
## Description ### 1. Why the change? Add Nano how-to guides to our documentation for better user experience. How-to guides aim at giving multiple bite-sized, task-oriented, and executable examples for...
## Description This PR makes is_predict=True works for BaseForecaster ### 1. Why the change? https://github.com/intel-analytics/BigDL/issues/5874 ### 2. User API changes Before API is : ```python # BaseForecaster x, y =...
API doc is missing for `bigdl.nano.tf.keras.layers.Embedding` and `bigdl.nano.tf.keras.layers.Embedding`
## Description Provide a How-to guide for new function: predict_interval in Forecaster ### 1. Why the change? Provide a How-to guide for new function: predict_interval in Forecaster ### 2. User...
## Description Bigdl-nano InferenceOptimizer example about how to export the optimized model to standard format. ### 1. Why the change? We need to show users how to export the optimized...
## Description Upgrade gramine SGX enclave to MREnclave, which is signed with hash values, and can be registered and attested by AS. ### 1. Why the change? Eliminate the leakage...
## Background Built-in dataset can be downloaded and preprocessed by `get_public_dataset`, but considering that users need dataset to benchmark, we may need to add a new API to help users...
## Description when call ```python orca_estimator = Estimator.from_torch(model=model_creator, optimizer=optimizer_creator, loss=criterion, metrics=[Accuracy()], model_dir=model_dir', use_tqdm=True, backend=backend) ``` on Databricks cluster, model_dir should be a shared path to save files, which is a...
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...