Comments (2)
Hello, you mentioned optimizations for scoring time in #4630
P1 (Large) Replace CPU-based batch expansion with multi-query attention kernel call
I think multi-query attention kernel is not equal to MQA here, it is more like the append stage in flashinfer, am I right?
And I notice that the calculation process of append is similar to that of chunked prefill's one step. So I use chunked prefill to implement the AppendTop1Scorer which get a 10% speedup compared to BatchExpansionTop1Scorer. It's a dirty solution, since I create a new SequenceGroupMetadata which change the scoring sequence to a chunked prefill sequence. This implementation conflicts with recompute and chunked prefille.
So the perfect implementation should be that ModelRunner and Backend support the append stage, Backend should already support it if it supports chunked prefill.
In addition, is this issue about solving the scheduling problem of speculative decoding? Can you give a detailed introduction to what needs to be done in this issue?
from vllm.
That's awesome. You should chat with @LiuXiaoxuanPKU who is removing batch expansion from vLLM.
FYI this issue is about combining the ITL improvements obtained from chunked prefill scheduling with spec decode.
from vllm.
Related Issues (20)
- stream_options.include_usage does not work HOT 1
- gfx908 architecture not working for version 0.5.1 HOT 3
- [Bug]: llava model gets stuck with RuntimeError: Please increase the max_chunk_bytes parameter. HOT 3
- [RFC]: A Graph Optimization System in vLLM using torch.compile HOT 2
- [Bug]: Runtime AssertionError: 32768 is not divisible by 3, multiproc_worker_utils.py:120, when using 3 GPUs for tensor-parallel HOT 6
- [Bug]: Problem loading Gemma 2 27b-it HOT 1
- [Bug]: Gemma-2 + FlashInfer: ValueError: Unsupported max_frags_z:
- [Misc]: _run_workers_async function of DistributedGPUExecutorAsync HOT 1
- when i set tensor_parallel_size>1(A100 * 4), it does not work HOT 8
- [Bug]: `samplers/test_logprobs.py` fail on H100
- [Bug]: Timeout Error When Deploying Llamafied InternLM2-5-7B-Chat-1M Model via vLLM OpenAI API Server
- [Feature]: Apply chat template through `LLM` class HOT 11
- [Bug]: When using qwen-32b-chat-awq with multi-threaded access, errors occur after approximately several hundred visits.”vllm.engine.async_llm_engine.AsyncEngineDeadError: Background loop has errored already.“ HOT 1
- [Feature]: Return softmax of attention layer. HOT 3
- [Bug]: Paligemma support for PNG files HOT 5
- [Bug]: illegal memory access when increase max_model_length on FP8 models HOT 6
- [Bug]: autogen can't work with vllm v0.5.1
- v0.5.2, v0.5.3, v0.6.0 Release Tracker HOT 5
- [Bug]: Severe computation errors when batching request for microsoft/Phi-3-mini-128k-instruct HOT 3
- [Bug]: The shape of the embed_tokens of llama model doesn't match the llama3 configuration
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from vllm.