### What this PR does / why we need it?
Earlier we fixed a similar issue for qwen2.5-vl 【
https://github.com/vllm-project/vllm-ascend/issues/4430 】, and then the
multimodal models in vllm v0.11.0 should all have this problem. Here, we
have specifically proposed a fix for qwen3-vl-moe.
---------
Signed-off-by: Levi-JQ <yujinqi2@huawei.com>
Co-authored-by: Levi-JQ <yujinqi2@huawei.com>
### What this PR does / why we need it?
#3985 move stream context initialization before for-loops to improve
performance. However, we find that this might cause potential accuracy
drop when used with pd disaggregation. Thus we partly revert this change
when using pd disaggregation, and we shall fix this bug in th future.
### Does this PR introduce _any_ user-facing change?
No.
---------
Signed-off-by: Angazenn <supperccell@163.com>
### What this PR does / why we need it?
Resolve the interface compatibility issue of get_input_embeddings in MM，
because the get_input_embeddings func of other model does not have the
is_multimodal parameter
---------
Signed-off-by: Levi-JQ <yujinqi2@huawei.com>
Co-authored-by: Levi-JQ <yujinqi2@huawei.com>
### What this PR does / why we need it?
Previously, the dummy run executed compute_logits only once, regardless
of num_speculative_tokens. This caused execute_model to hang on
compute_logits when lm head tensor parallelism exceeded 1. The fix
ensures compute_logits executes correctly during dummy run, matching
num_speculative_tokens.
Signed-off-by: zouyida2052 <zouyida2002@gmail.com>
<!-- Thanks for sending a pull request!
BEFORE SUBMITTING, PLEASE READ
https://docs.vllm.ai/en/latest/contributing/overview.html
-->
### What this PR does / why we need it?
disable NZ for float weight case. This is only a quick fix for dev
branch.
For main branch, we'll consider more case to make it more common.
### Does this PR introduce _any_ user-facing change?
<!--
Note that it means *any* user-facing change including all aspects such
as API, interface or other behavior changes.
Documentation-only updates are not considered user-facing changes.
-->
### How was this patch tested?
qwen2.5 32B
<img width="441" height="221" alt="image"
src="https://github.com/user-attachments/assets/7ae18ffd-1ce2-43d9-9960-be45250ad0da"
/>
---------
Signed-off-by: 刘哲续 <liuzhexu1@huawei.com>
Co-authored-by: 刘哲续 <liuzhexu1@huawei.com>
### What this PR does / why we need it?
When cudagraph_mode is set to FULL_DECODE_ONLY, if dp > 1, the dummy-run
process will be triggered. When calling the update_attn_params function,
the num_tokens parameter needs to be passed, and this value is obtained
through positions.shape[0]. However, the multimodal model uses mRope
(multi-dimensional rotary positional embeddings), which causes the shape
of positions to be 2. As a result, the value obtained from
positions.shape[0] is incorrect. We solve this problem by replacing
positions.shape[0] with num_tokens.
### Does this PR introduce _any_ user-facing change?
NO
### How was this patch tested?
vLLM version: v0.11.0rc3
vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: wujinyuan1 <wjy9595@qq.com>
Co-authored-by: wujinyuan1 <wjy9595@qq.com>
<!-- Thanks for sending a pull request!
BEFORE SUBMITTING, PLEASE READ
https://docs.vllm.ai/en/latest/contributing/overview.html
-->
### What this PR does / why we need it?
This is cherry-pick from #4097 .
Currently, we set `seq_lens` in dummy attn_metadata to be
`max_model_len` to get max workspace for attention during capturing.
However, setting it consistently to be `max_model_len` causing dummy_run
to execute a long attention when running actual inference. For example,
if there is a single req with `seqs_lens` as [8] but `max_model_len` is
131072, the whole process will be slow down by dummy_run as it execute a
fake long-seq attention. Therefore, we instead set it to max_query_len,
which is also consistent with vLLM gpu implementation.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
<!--
CI passed with new added/existing test.
If it was tested in a way different from regular unit tests, please
clarify how you tested step by step, ideally copy and paste-able, so
that other reviewers can test and check, and descendants can verify in
the future.
If tests were not added, please describe why they were not added and/or
why it was difficult to add.
-->
---------
Signed-off-by: Angazenn <supperccell@163.com>
This PR upgrade CANN from 8.2rc1 to 8.3rc1 and remove the CANN version
check logic.
TODO: we notice that UT runs failed with CANN 8.3 image. So the base
image for UT is still 8.2. We'll fix it later.
- vLLM version: v0.11.0
- vLLM main:
83f478bb19
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
### What this PR does / why we need it?
Fix the issue of MTP being enabled and setting
Imhead_tensor_parallel_size=16 causing the inference to hang.
Signed-off-by: wyh145 <1987244901@qq.com>
### What this PR does / why we need it?
1. Revert [bugfix for mtp in
fullgraph](0948483642)
and support it when vllm supports
2. raise error when cudagraph_capture_sizes can't be an integer multiple
of uniform_decode_query_len
3. bugfix when max_num_seqs=14 in mtp=2 scenario
---------
Signed-off-by: zouyida2052 <zouyida2002@gmail.com>
### What this PR does / why we need it?
bugfix for mtp in fullgraph
### Does this PR introduce _any_ user-facing change?
no
---------
Signed-off-by: zouyida2052 <zouyida2002@gmail.com>
### What this PR does / why we need it?
force with_prefill true after allreduce in kv producer. This is a backport of #3768 and #3849
---------
Signed-off-by: liziyu <liziyu16@huawei.com>
<!-- Thanks for sending a pull request!
BEFORE SUBMITTING, PLEASE READ
https://docs.vllm.ai/en/latest/contributing/overview.html
-->
cherry pick https://github.com/vllm-project/vllm-ascend/pull/3677
Remove redundant operations from `model_runner` and `forward_context`.
This optimization can significantly reduce the idle time (bubble) before
decoding when running models with small parameter counts (e.g.,
Qwen/Qwen2.5-0.5B).
Testing on 800I A2, bubble is reduced from 3.8ms to 2.8ms :
Before
<img width="1655" height="696" alt="image"
src="https://github.com/user-attachments/assets/d7608e52-2438-46dd-8fc9-391fd6274495"
/>
After
<img width="1607" height="774" alt="image"
src="https://github.com/user-attachments/assets/56daf081-2dba-4d2e-99d4-e055187d9806"
/>
### What this PR does / why we need it?
<!--
- Please clarify what changes you are proposing. The purpose of this
section is to outline the changes and how this PR fixes the issue.
If possible, please consider writing useful notes for better and faster
reviews in your PR.
- Please clarify why the changes are needed. For instance, the use case
and bug description.
- Fixes #
-->
### Does this PR introduce _any_ user-facing change?
<!--
Note that it means *any* user-facing change including all aspects such
as API, interface or other behavior changes.
Documentation-only updates are not considered user-facing changes.
-->
No
### How was this patch tested?
<!--
CI passed with new added/existing test.
If it was tested in a way different from regular unit tests, please
clarify how you tested step by step, ideally copy and paste-able, so
that other reviewers can test and check, and descendants can verify in
the future.
If tests were not added, please describe why they were not added and/or
why it was difficult to add.
-->
---------
Signed-off-by: realliujiaxu <realliujiaxu@163.com>
### What this PR does / why we need it?
Caps the calculated maximum number of tokens at 512.
This prevents allocating an excessively large buffer when a cudagraph
capture size is not specified, mitigating the risk of out-of-memory
errors.
### Does this PR introduce _any_ user-facing change?
None.
### How was this patch tested?
None.
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
This is a port PR of #3636 .
Move the creation of dummy attention metadata to occur after the ACL
graph runtime mode is determined. This ensures the metadata is
initialized with the correct configuration during a profile run.
Additionally, remove the `attn_metadata` existence check before updating
MLA attention parameters. This change prevents the update from being
skipped when metadata is not yet available, ensuring parameters are set
correctly.
### Does this PR introduce _any_ user-facing change? None.
### How was this patch tested?
None.
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
1.Add eplb ci to check the change of eplb feature.
2.Add param checking of eplb params.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
Qwen in A3.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: offline0806 <3337230449@qq.com>
Co-authored-by: offline0806 <3337230449@qq.com>
### What this PR does / why we need it?
Fix mtp tochair bug cuased by #2719
Since FIA need extra space for padding, we need to enforce
`self.max_num_seqs > self.scheduler_config.max_num_seqs` in KV consumer
+ MTP
This means that, `self.max_num_seqs` **>** the actual maximum requests
(`self.scheduler_config.max_num_seqs`)
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: xuyexiong <xuyexiong@huawei.com>
### What this PR does / why we need it?
The `force_attention` parameter is designed for flash infer kernel
warmup, we don't actually need it on Ascend device (at least for
now).And it tends to make things more complicated. So we replace the
`force_attention` parameter with `aclgraph_runtime_mode` in the
attention metadata creation logic.
This change makes the control flow more explicit by directly using the
graph runtime mode to determine how to build attention metadata, rather
than relying on an intermediate boolean flag. This simplification
removes redundant logic and clarifies the conditions for building
attention metadata for full decode graph mode.
### Does this PR introduce _any_ user-facing change?
None.
### How was this patch tested?
DP + `FULL_DECODE_ONLY` + online serving.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
Update the check logic of reinitializing inputbatch, this is a follow-up
pr of #3477. `kernel_block_sizes` is a `list[list[int]]` and the
original logic will always update `InputBatch` when using hybrid blocks,
this pr fixes that
### How was this patch tested?
locally test with qwen3-next
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: MengqingCao <cmq0113@163.com>
### What this PR does / why we need it?
Fix attn_group initialization timing so that fix qwen3-next model
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: MengqingCao <cmq0113@163.com>
### What this PR does / why we need it?
Add new npu_fused_infer_attention_score op to improve perfomance in
flash attention case.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: ZYang6263 <zy626375@gmail.com>
### What this PR does / why we need it?
In memory of https://github.com/vllm-project/vllm-ascend/pull/2610 and
#3449 Fix Mtp torchair pd bug.
In the pd Disaggregation scenario, the first token of the inference
after the d node receives the kv follows the eager mode.
Fixes:
Running with MTP torchair graph mode with Prefilling Decoding
Disaggregation , if all requests processed by the D node are requests
just transmitted from the P node, it will break the torchair graph.
Reason: During PD Disaggregation , the P node only transmits the KV
cache and prompt to the D node, not the actual tokens inferred (neither
the main model tokens nor the MTP tokens are transmitted). Therefore,
the D node will treat this request as one without MTP tokens for
inference (seq_len=1).
The community does not have graph mode issues because the community's
attention has a seq_len=1 for each batch during the decode phase.
We have issues because the graph mode pads according to processing 2
tokens per request. When there are some seq_len=1 and some seq_len=2,
padding is done at the end. If all requests received by the D node are
seq_len=1, padding cannot be performed normally according to the
attention's fia operator constraints.
Solution:
The kv consumer uses extra torchair graph padding to avoid breaking FIA
graph constrains (The one this PR implemented).
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: xuyexiong <xuyexiong@huawei.com>
### What this PR does / why we need it?
This PR aims to add padding logic to seq_lens、block_tables when running
in full decode scenario. Before this PR, the number of input tokens with
padding might exceeds corresponding seq_lens. For example, when running
in full decode scenario:
```
input_ids : [1, 3, 0, 0]
seq_lens: [2, 1]
query_start_loc: [0, 1, 2]
```
Here, `input_ids` is padded by 2 tokens while
`seq_lens`/`query_start_loc` are not. The mismatch between `input_ids`
and `seq_lens`/`query_start_loc` might cause some potential bugs. This
PR would change it into :
```
input_ids : [1, 3, 0, 0]
seq_lens: [2, 1, 1, 1]
query_start_loc: [0, 1, 2, 3, 4]
```
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: Angazenn <supperccell@163.com>
### What this PR does / why we need it?
Fix 3 bugs in flash comm1 of Allgather
EP(https://github.com/vllm-project/vllm-ascend/pull/3334):
1. call `enable_sp()` with argument `vllm_config` trigger a lot of
warning log, this PR caches its return value.
2. `num_tokens_after_padding` should be cpu tensor as it will used as
`num_tokens_across_dp_cpu` in `DPMetadata`. It will causes may d2h copy
when running model.
3. In PD, model runner will execute `kv_connector_no_forward`，where
`num_tokens` is None
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: realliujiaxu <realliujiaxu@163.com>
### What this PR does / why we need it?
Make the Full Graph mode can run with MTP.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: anon189Ty <Stari_Falcon@outlook.com>
### What this PR does / why we need it?
Currently, MTP Model in deepseek can not be capture in ACLGraph. This PR
is use to allow MTP to be captured in ACLGraph mode.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: anon189Ty <Stari_Falcon@outlook.com>
### What this PR does / why we need it?
MLA and GQA use different computation logic: MLA slice batches and only
compute on the actually valid tokens. That means outer padding must be
handled carefully — the accuracy issue this PR fixes was caused by stale
data in `slot_mapping` being reused by subsequent inference steps.
So we zeros out the portion of the slot mapping tensor that is not used
by the currently scheduled tokens.
### Does this PR introduce _any_ user-facing change?
None.
### How was this patch tested?
Working on it.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
Adapt deepseek-v3.2 to vllm 0.11.0, removing the useless patch.
The final goal is to remove all the patches and align the code arch to
vllm, thus we need to do the following work in next prs.
TODO:
- [x] remove patch on attention spec
- [ ] refactor the kvcache creation logic
### Does this PR introduce _any_ user-facing change?
N/A
### How was this patch tested?
1. CI passed with existing test.
2. Test pass with deepseek-v3.2-exp
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: MengqingCao <cmq0113@163.com>
What this PR does / why we need it?
1.Record expert map without dynamic eplb.
2.Add export PYTHONOPTIMIZE=1 when using dynamic eplb.
3.change eplb doc
Does this PR introduce any user-facing change?
How was this patch tested?
Qwen3_moe in A3.
- vLLM version: v0.11.0
---------
Signed-off-by: offline0806 <3337230449@qq.com>
Co-authored-by: offline0806 <3337230449@qq.com>
### What this PR does / why we need it?
[Feat] Supports Aclgraph for bge-m3
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
```
pytest -s tests/e2e/singlecard/test_embedding.py
pytest -s tests/e2e/singlecard/test_embedding_aclgraph.py
```
to start an online server with bs 10, each batch's seq length=8192, we
set --max-num-batched-tokens=8192*10 to ensure encoder is not chunked:
```
vllm serve /home/data/bge-m3 --max_model_len 1024 --served-model-name "bge-m3" --task embed --host 0.0.0.0 --port 9095 --max-num-batched-tokens 81920 --compilation-config '{"cudagraph_capture_sizes":[8192, 10240, 20480, 40960, 81920]}'
```
For bs10, each batch's seq length=8192, QPS is improved from 85 to 104,
which is a 22% improvement, lots of host bound is reduced.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: xuyexiong <xuyexiong@huawei.com>
Co-authored-by: wangyongjun <1104133197@qq.com>
### What this PR does / why we need it?
Because the update_attn_params function requires passing the num_tokens
parameter, and num_tokens is obtained via postions.shape[0]. However,
the multimodal model uses mrope (Multidimensional Rotary Position
Embedding), which results in the postions having a shape of 2.
Consequently, postions.shape[0] retrieves an incorrect value.We resolve
this issue by replacing positions.shape[0] with maybe_padded_num_tokens.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: fan2956 <zhoufan53@huawei.com>
### What this PR does / why we need it?
Refactor KVCache as page_size_bytes is ineffective.
1. Currently the `AttentionSpec` is patched, but the `page_size_bytes`
is still using that in vLLM in runtime, thus the patch is not working
actually. Thus this pr removes the patch on `AttentionSpec`, and will do
the final fix in vLLM.
2. Use `MLAAttentionSpec` instead of `FullAttentionSpec` to reduce
`page_size_bytes` of spec, so that num_blocks in spec could double
### How was this patch tested?
Test pass with Qwen3-Next and DeepSeek-V3.2-Exp
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: MengqingCao <cmq0113@163.com>
### What this PR does / why we need it?
Currently, when executing to the Linear layer of models in vLLM-Ascend,
the weights format is ND in unquantized case and skipped ascend case.
This PR supplements the execution logic for Linear layer. We use a new
global variable: VLLM_ASCEND_ENABLE_NZ. When VLLM_ASCEND_ENABLE_NZ=1 and
CANN version is 8.3, the weights of the Linear layer will be converted
to FRACTAL_NZ, in both unquantized case and skipped ascend case. We also
use VLLM_ASCEND_ENABLE_NZ to control the existing NZ conversion, such as
w8a8-quantized case.
### Does this PR introduce _any_ user-facing change?
Add a new global variable VLLM_ASCEND_ENABLE_NZ. If you want to use NZ
format, you should set VLLM_ASCEND_ENABLE_NZ=1.
### How was this patch tested?
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
Signed-off-by: anon189Ty <Stari_Falcon@outlook.com>
### What this PR does / why we need it?
Replaces the hardcoded `mc2_tokens_capacity` with the max graph capture
size for a more accurate allocation.
This change ensures the capacity is correctly sized relative to the
graph capture configuration, removing a magic number and making the
setup more robust.
This PR fixes two issues:
1. <del>MC2 op restrictions differ between SoCs.</del> @Angazenn This
requires an overhaul, hence removed from this PR, please commit another
PR.
2. The hardcoded value `512` allocates too much buffer for large models.
### Does this PR introduce _any_ user-facing change?
None.
### How was this patch tested?
Tested in daily checks.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
This PR primarily focuses on two key changes:
1. Adjusts internal interface calls to optimize the interaction logic
between related modules.
2. Exposes an interface that allows users to select the EPLB algorithm,
enabling more flexible configuration based on specific usage scenarios.
These changes aim to enhance the usability of the system while ensuring
the stability of internal operations. Relevant unit tests have been
updated to cover the modified logic.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: Che Ruan <cr623@ic.ac.uk>
Co-authored-by: Che Ruan <cr623@ic.ac.uk>
### What this PR does / why we need it?
Adds support for capturing the Multi-Layer Attention (MLA) decode
operation into an ACL graph. This improves performance by compiling the
attention kernel for single-token decoding.
Key changes include:
- Implementing the graph capture logic for the MLA kernel, including
workspace management and parameter updates.
- Modifying the rotary embedding (RoPE) handling to use pre-allocated
tensors, which is a requirement for graph capture.
- Adding a `build_for_graph_capture` method to the MLA metadata builder
to create dummy metadata during the graph compilation phase.
Known issues:
- Currently, MTP is not supported in FULL_DECEDE_ONLY mode -- we're
working on a fix
- We are preparing to remove update_mla_attn_params with
auto_dispatch_capture
### Does this PR introduce _any_ user-facing change?
compilation_config={
"cudagraph_mode": "FULL_DECODE_ONLY",
},
### How was this patch tested?
- vLLM version: v0.11.0
---------
Signed-off-by: panchao-hub <315134829@qq.com>
Signed-off-by: p00465316 <panchao13@huawei.com>
Co-authored-by: p00465316 <panchao13@huawei.com>
Co-authored-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
- Refacotr and integrate a unified `WeightPrefetchMethod`
- Integrate `qkv_proj.weight` and `o_proj.weight` in quantized Attention
modules
- Prefetching these weights ahead of matmul-like operators imporves
performance by reducing L2 cache transfer latency
### Does this PR introduce _any_ user-facing change?
Add a new config in `--additional-config` for configuration:
```json
{
"weight_prefetch_config": {
"enabled": false,
"prefetch_ratio": {
"attn": {
"qkv": 1.0,
"o": 1.0,
},
},
},
}
```
This feature is enabled by default, and can be disabled through this
configuration
### How was this patch tested?
- vLLM version: v0.11.0
---------
Signed-off-by: yuzhup <15705211260@163.com>
Signed-off-by: zhoux77899 <zhouxiang100@huawei.com>
Co-authored-by: yuzhup <15705211260@163.com>
### What this PR does / why we need it?
When running DP in a non-equilibrium scenario, which means there is some
dp groups executing `dummy_run`, we need to make sure it running the
same mode as other dp, thus improving then performance in dp scenario
### How was this patch tested?
Tested by adding log in `_dummy_run`
- vLLM version: v0.10.2
- vLLM main:
https://github.com/vllm-project/vllm/commit/releases/v0.11.0
---------
Signed-off-by: MengqingCao <cmq0113@163.com>
### What this PR does / why we need it?
Currently we run an extra profile_run with `num_tokens ==
self.mc2_tokens_capacity`. However, when setting `max_num_batched_tokens
< self.mc2_tokens_capacity`, this will trigger an assertion error that
requires num_tokens in `_dummy_run` to be smaller than
`max_num_batched_tokens`. This PR skips this extra `profile_run` if
`self.max_num_tokens <= self.mc2_tokens_capacity` so as to avoid this
bug.
This PR fixes a bug that `kernel_block_sizes` never equals to
`[self.cache_config.block_size]`. `kernel_block_sizes` is type of
List[List[int]], so the condition should be `kernel_block_sizes !=
[[self.cache_config.block_size]]`. This also helps to resolve a issue
that cpu_offload_gb cannot be enabled.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
- vLLM version: v0.10.2
- vLLM main:
https://github.com/vllm-project/vllm/commit/releases/v0.11.0
Signed-off-by: Angazenn <supperccell@163.com>
### What this PR does / why we need it?
Fix the error "cur batch_size is invalid" during profile_run in the
torchair scenario.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.10.2
- vLLM main:
https://github.com/vllm-project/vllm/commit/releases/v0.11.0
Signed-off-by: WithHades <244036962@qq.com>
### What this PR does / why we need it?
1. Solved the issue where sizes capture failed for the Qwen3-32b-int8
model when aclgraph, dp1, and tp4 were enabled.
2. Added the exception thrown when sizes capture fails and provided a
solution
3. Add this common problem to the FAQ doc
### Does this PR introduce _any_ user-facing change?
no
### How was this patch tested?
ut
- vLLM version: v0.10.2
- vLLM main:
https://github.com/vllm-project/vllm/commit/releases/v0.11.0
Signed-off-by: lilinsiman <lilinsiman@gmail.com>
### What this PR does / why we need it?
fix oom in aclgraph.
1. In the current token dispatch implementation, tensors are mounted on
class instances to facilitate parameter passing between different
methods. This approach prevents automatic recycling of these tensors. In
some cases, it may lead to out-of-memory error. To address this issue,
we manually set these tensors to None to release corresponding memory.
2. The `profile_run` method is designed to accurately estimate the
maximum NPU memory usage during vLLM inference. However, in certain
scenarios, MoE models perform inference via MC2, which includes
communication and consumes additional NPU memory. This leads to
inaccurate estimation by the profile run. We address this by actively
triggering the MC2 during profile run for initialization.```.
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
- vLLM version: v0.10.2
- vLLM main:
52d0cb8458
Signed-off-by: WithHades <244036962@qq.com>
