1.17 documentation update #172
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…a/multiprocessing_hpu
This reverts commit 1546cd5.
| `HPU Graphs <https://docs.habana.ai/en/latest/PyTorch/Inference_on_PyTorch/Inference_Using_HPU_Graphs.html>`__ are currently the most performant execution method of vLLM on Intel Gaudi. When HPU Graphs are enabled, execution graphs will be traced (recorded) ahead of time (after performing warmup), to be later replayed during inference, significantly reducing host overheads. Recording can take large amounts of memory, which needs to be taken into account when allocating KV cache. Enabling HPU Graphs will impact the number of available KV cache blocks, but vLLM provides user-configurable variables to control memory management. | ||
|
|
||
|
|
||
| Whenever HPU Graphs are being used, they share the common memory pool ("usable memory") as KV cache, determined by ``gpu_memory_utilization`` flag (``0.9`` by default). Environment variable ``VLLM_GRAPH_RESERVED_MEM`` defines the ratio of memory reserved for HPU Graphs capture. With its default value (``VLLM_GRAPH_RESERVED_MEM=0.4``), 40% of usable memory will be reserved for graph capture (later referred to as "usable graph memory"), and the remaining 60% will be utilized for KV cache. Before KV cache gets allocated, model weights are loaded onto the device, and a forward pass of the model is executed on dummy data, to estimate memory usage. Next, KV cache gets allocated, model is warmed up, and HPU Graphs are captured. Environment variable ``VLLM_GRAPH_PROMPT_RATIO`` determines the ratio of usable graph memory reserved for prefill and decode graphs. By default (``VLLM_GRAPH_PROMPT_RATIO=0.5``), both stages have equal memory constraints. Lower value corresponds to less usable graph memory reserved for prefill stage, e.g. ``VLLM_GRAPH_PROMPT_RATIO=0.2`` will reserve 20% of usable graph memory for prefill graphs, and 80% of usable graph memory for decode graphs. |
There was a problem hiding this comment.
I would consider to rephrase this paragraph and present it as table or so. You can use following columns, e.g.: env_name, default, available memory, used_memory, free_memory, description for each env.
There was a problem hiding this comment.
I'm not sure if it makes sense to write this down as a table - this section describes step-by-step how the memory allocations look like. The documentation for each env var is in the section below, if you want just that. The goal here is to describe the design on some example.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.