Comments (14)
Does the sequence length plus proposal length go over the max model length ?
That was our suspicion as well so we made speculative-max-model-len shorter than the max-model-len - num-speculative-tokens but that doesnt seem to stop that issue.
--max-model-len 16384 \
--speculative-max-model-len 16000 \
--speculative-model [ngram] \
--num-speculative-tokens 128 \
--ngram-prompt-lookup-max 32 \
--ngram-prompt-lookup-min 16 \
from vllm.
yep.
from vllm.
@njhill we ran into this error with 0.4.3 originally before we tried upgrading to 0.5.0.
from vllm.
Thanks for creating the issue. Two questions:
- Does the problem still occur if prefix caching is disabled?
- Does the problem still occur if cuda graphs are disabled?
from vllm.
@cadedaniel Thanks for the quick response!
- I tried without
--enable-prefix-caching
and it eventually ran into the same error. - Then I tried without
--enable-prefix-caching
, and enabled--enforce-eager
. This didn't error on the set of queries I ran.
from vllm.
Thanks for trying those out so fast :)
OK the issue is very likely caused by CUDA graphs + batch expansion. This should be fixed, but currently since spec decode performance isn't good, it won't be prioritized until after that.
FYI @LiuXiaoxuanPKU another issue with batch expansion + cuda graph
from vllm.
Do you recommend just using --enforce-eager
until this is fixed?
from vllm.
If you are blocked by this issue, the fix shouldn't be very hard. I think we simply need to configure the cuda graph max size to include the expanded batch size.
from vllm.
The code that is breaking is:
if use_captured_graph:
# The shape of graph_block_tables is
# [max batch size, max context len // block size].
input_block_tables = self.graph_block_tables[:batch_size]
for i, block_table in enumerate(block_tables):
if block_table:
input_block_tables[i, :len(block_table)] = block_table
The issue is that len(block_table) > input_block_tables.shape[1]
and the second dimension corresponds to max context len // block size
. Am i misunderstanding in how this is a batch-size issue and not a context len issue?
from vllm.
good point. Wonder why this is specific to spec decode then.
Does the sequence length plus proposal length go over the max model length ?
from vllm.
@Adhyyan1252 could you see if you also get this error with vLLM 0.4.3?
from vllm.
try to add params : '--max-seq-len-to-capture' eqauls to max_model_len
from vllm.
Still the same issue
from vllm.
Still the same issue..
from vllm.
Related Issues (20)
- [Feature]: Please Support FATRelu
- [Usage, bug]: vLLM Docker | ValueError: OpenTelemetry packages must be installed before configuring 'otlp_traces_endpoint' during vLLM startup
- [Performance]: vllm量化和非量化的性能对比 HOT 3
- [Bug]: when `echo=True`, vllm will append chat template(`assistant`) after the last message HOT 6
- [Misc]: how to add tests for new backends? HOT 1
- [Feature]: Pipeline Parallelism support for the Vision Language Models HOT 4
- [Usage]: Qwen2 GGUF model can't run successfully HOT 7
- [Feature]: Overlap model weight loading and model prefill HOT 2
- [RFC]: Add Ascend NPU as a new backend
- [Bug]: Inconsistent Output Behavior with and without tools and tool_choice Parameters
- [Usage]: Periodic snapshots for spot instances HOT 1
- [RFC]: Enable Memory Tiering for vLLM HOT 5
- [Bug]: vLLM inconsistently crashes on startup for multinode cluster HOT 7
- [Usage]: how to abort request? HOT 1
- [Usage]: alignment between trl and llm.generate
- [Bug]: errors when loading mixtral 8x7b
- [Bug]: The MixtralForCausalLM architecture and the mistralai/Mixtral-8x7B-Instruct-v0.1 model are stated to be supported by vLLM, but an error occurs during model loading. HOT 3
- [Bug]: Unable to use fp8 kv cache with neuralmagic quants on ampere HOT 1
- [Doc]: AutoAWQ quantization example fails HOT 3
- [Bug]: Error loading microsoft/Phi-3.5-vision-instruct HOT 1
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.