Kevin-shihello-world issues

Results 4 issues of


                                            Kevin-shihello-world

I wonder if we can use batch inference and offload in mii pipeline ?

Since in the principle of the latest fastgen in mii includes Continuous Batching, We should have choice to adjust the batch size in mii pipeline But haven't found it in...

Why the FeatureExpander function got wrong to init?

I want to do some tests on the GFN code. And I follow the lead but I want to just run on one dataset so I modified the run_exp_benchmark function...

How to let diffusers use local code for pipelineinstead of download it online everytime We use it?

I tried to use the instaflowpipeline from example/community to.run my test However, even after i git cloned the repository to my environment it still Keep trying to Download the latest...

stale

With accelerate already well installed itstill show this in utils/registry.py:LlamaForTextGeneration: Using `low_cpu_mem_usage=True` or a `device_map` requires Accelerate: `pip install accelerate`

I tried to run llama according to the best implement in :https://www.modelscope.cn/headlines/article/170, however the utils/registry.py code keeps warning me LlamaForTextGeneration: Using `low_cpu_mem_usage=True` or a `device_map` requires Accelerate: `pip install accelerate`...