### What this PR does / why we need it?
- Fixes Qwen3-Next enable nz accuracy problem
---------
Signed-off-by: wxsIcey <1790571317@qq.com>
Signed-off-by: Icey <1790571317@qq.com>
### What this PR does / why we need it?
cherry-pick from https://github.com/vllm-project/vllm-ascend/pull/4022
The code bug caused an empty bubble. When the npu_paged_cache_load
operator was called, it forcibly transferred seq_len2 to the device,
which triggered synchronization and interrupted the CPU operator's
launch stream.
---------
Signed-off-by: underfituu <hzhucong@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?
Remove test cases containing CPU incompatible operators
### 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: wangli <wangli858794774@gmail.com>
This PR deletes model codes of deepseek_v2 and deepseek_v3 to reuse the
model file from vLLM.
vLLM Ascend now uses custom ops register way instead of model file
hard-coding.
- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0
---------
Signed-off-by: whx-sjtu <2952154980@qq.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?
Calculate in advance the workspace memory size needed for the
PagedAttention operator to avoid deadlocks during resource cleanup. This
PR requires torch_npu version 0920 or newer.
### 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: wangxiaoxin-sherie <wangxiaoxin7@huawei.com>
Co-authored-by: wangxiaoxin-sherie <wangxiaoxin7@huawei.com>
### 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>
we notice that torch npu 0919 doesn't work. This PR revert related
change which rely on 0919 version.
Revert PR: #3295#3205#3102
Related: #3353
- vLLM version: v0.11.0
### What this PR does / why we need it?
Calculate in advance the workspace memory size needed for the
PagedAttention operator to avoid deadlocks during resource cleanup. This
PR requires torch_npu version 0920 or newer.
### How was this patch tested?
- vLLM version: v0.11.0
---------
Signed-off-by: wangxiaoxin-sherie <wangxiaoxin7@huawei.com>
Co-authored-by: wangxiaoxin-sherie <wangxiaoxin7@huawei.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>
Upgrade vLLM to newest commit.
1. Remove the useless func get_state_cls, it has been removed from vLLM
already.
e6750d0b18
2. Fix ut broken by
6160ba4151
- vLLM version: v0.10.2
- vLLM main:
b1068903fd
---------
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Note: This depends on [vLLM
#25161](https://github.com/vllm-project/vllm/pull/25161) and the
torch\_npu release from September 30.
### What this PR does / why we need it?
This pull request adds `FULL_DECODE_ONLY` mode for GQA/MHA models (MLA
models like DeepSeek V3/R1 are not included). Key improvements include:
* **Reduced dispatch latency:** By replaying the entire model execution
graph at once, we cut overhead compared with multiple smaller replays.
* **Stabilized multi-device performance:** Captureing the whole model as
one static graph also mitigates the dispatch fluctuations across
devices.
* **Stream/resource savings:** Consolidating graph captures frees up
streams, allowing more graphs to be captured.
**Known issues:**
1. `_npu_paged_attention` currently manages its own workspace in
`torch_npu`, which can deadlock when synchronizing during graph replay —
we’re working on a fix.
There may be other corner cases. This PR is the first in a planned
series; we’ll continue to iterate and address remaining issues in
follow-ups.
This is essentially a port of #1503 and #1677, but includes two major
changes:
1. Let `graph_dispatcher` decide the graph mode instead of hard-coding
it in the backend, which decouples Full Graph and Piecewise Graph and
could make it possible to remove dynamo.
2. Adapt to the new `attn_group` logic, but leave a small hack in
`update_graph_params`; multi-attention models may or may not be fully
supported yet.
### Does this PR introduce _any_ user-facing change?
```python
compilation_config={
"cudagraph_mode": "FULL_DECODE_ONLY",
},
```
### How was this patch tested?
Tests included.
- vLLM version: v0.10.2
- vLLM main:
9607d5eb44
---------
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
### What this PR does / why we need it?
Fix shape not match when test LLM-Research/Phi-4-mini-instruct accuarcy
### Does this PR introduce _any_ user-facing change?
Users can't set dynamic batch_size or use lm_eval test accuracy when
using models(sliding_window)
### How was this patch tested?
accuarcy of LLM-Research/Phi-4-mini-instruct is ok :
```
vllm (pretrained=LLM-Research/Phi-4-mini-instruct,max_model_len=4096,dtype=auto,tensor_parallel_size=1), gen_kwargs: (None), limit: None, num_fewshot: 5, batch_size: auto
|Tasks|Version| Filter |n-shot| Metric | |Value | |Stderr|
|-----|------:|----------------|-----:|-----------|---|-----:|---|-----:|
|gsm8k| 3|flexible-extract| 5|exact_match|↑ |0.8105|± |0.0108|
| | |strict-match | 5|exact_match|↑ |0.8097|± |0.0108|
```
- vLLM version: v0.10.2
- vLLM main:
3c96e7b8a1
Signed-off-by: hfadzxy <starmoon_zhang@163.com>
### What this PR does / why we need it?
This PR depends on the merge of #2707 and has adapted the aclgraph
functionality to support MTP.
### How was this patch tested?
- vLLM version: v0.10.2
- vLLM main:
2b85697031
---------
Signed-off-by: xuyexiong <xuyexiong@huawei.com>
### What this PR does / why we need it?
fix prefill attention bug，not support sliding window.
npu_fused_infer_attention_score head_dim only equal 128, not support
other number.
### Does this PR introduce _any_ user-facing change?
remove prefill phase npu_fused_infer_attention_score
### How was this patch tested?
- vLLM version: v0.10.1.1
- vLLM main:
e599e2c65e
---------
Signed-off-by: nsdie <yeyifan@huawei.com>
This PR fix a bug related to attention mask used in ring mla. Current
ring mla has supported compressed mask, so we can directly use a 512 *
512 attention mask.
- vLLM version: v0.10.1.1
- vLLM main:
b5ee1e3261
---------
Signed-off-by: whx-sjtu <2952154980@qq.com>
### What this PR does / why we need it?
Update UT for MLA case
- vLLM version: v0.10.1.1
- vLLM main:
14b4326b94
---------
Signed-off-by: SunnyLee219 <3294305115@qq.com>
### What this PR does / why we need it?
Add a sliding window size parameter to attention
### Does this PR introduce _any_ user-facing change?
### How was this patch tested?
Regarding the `Gemma3` model, set
additional_config={"ascend_scheduler_config": {"enabled":True}}, only
support AscendScheduler
test commond：`python3 -m vllm.entrypoints.openai.api_server --model
gemma3 --additional-config
'{"ascend_scheduler_config":{"enabled":true}}'`
- vLLM version: v0.10.1.1
- vLLM main:
6578e87365
---------
Signed-off-by: nsdie <yeyifan@huawei.com>
### What this PR does / why we need it?
This PR fix bugs and refactor cached mask generation logic. Now just
pre-construct and use the cached mask on cpu instead of device on npu.
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
CI passed with new added/existing test.
- vLLM version: v0.10.1.1
- vLLM main:
9b5f64238f
Signed-off-by: rjg-lyh <1318825571@qq.com>
### What this PR does / why we need it?
This PR move current unified mla backend to torchair folder and remove
torchair-related code in attention/mla_v1.py (1.3k -> 0.9k).
### Does this PR introduce _any_ user-facing change?
No.
### How was this patch tested?
Running eager mode with mla backend, and torchair mode with code before
[2445](https://github.com/vllm-project/vllm-ascend/pull/2445)
- vLLM version: v0.10.0
- vLLM main:
f571ff8eb6
Signed-off-by: linfeng-yuan <1102311262@qq.com>
### What this PR does / why we need it?
Fix some ci issue and refactor modelrunner
### Does this PR introduce _any_ user-facing change?
N/A
### How was this patch tested?
CI passed with existing test.
- vLLM version: v0.10.0
- vLLM main:
4d9c61993a
---------
Signed-off-by: wangli <wangli858794774@gmail.com>
Signed-off-by: MengqingCao <cmq0113@163.com>
Signed-off-by: weiguihua2 <weiguihua2@huawei.com>
Co-authored-by: wangli <wangli858794774@gmail.com>
Co-authored-by: weiguihua2 <weiguihua2@huawei.com>
### What this PR does / why we need it?
1.Shared Expert Sharding Strategy Update: Switched from TP-aligned to
pure DP for shared experts, enabling more efficient execution.
2.O_Proj AllReduce → ReduceScatter: Reduced communication overhead by
using ReduceScatter, made possible by pure DP sharding.
3.AllGather Postponed: Delayed to after QKV down projection to reduce
synchronization impact during prefill.
### How was this patch tested?
Adding ut case in `tests/ut/attention/test_mla_v1.py`
#### How to run
use parameter `--additional_config='{"enable_shared_expert_dp": true}'`
##### a.How to run eager mode
eg:
python -m vllm.entrypoints.openai.api_server --model=/model_path
--trust-remote-code -tp 8 -dp 2 --enable_expert_parallel --port 8002
--max-model-len 5120 --max-num-batched-tokens 16384 --enforce-eager
--disable-log-requests
--additional_config='{"ascend_scheduler_config":{"enabled":true},"enable_shared_expert_dp":
true,"chunked_prefill_for_mla":true}'
##### b.How to run graph mode
eg:
python -m vllm.entrypoints.openai.api_server --model=/model_path
--trust-remote-code -tp 8 -dp 2 --enable_expert_parallel --port 8002
--max-model-len 5120 --max-num-batched-tokens 16384
--disable-log-requests
--additional_config='{"ascend_scheduler_config":{"enabled":true},"enable_shared_expert_dp":
true,"chunked_prefill_for_mla":true,"torchair_graph_config":{"enabled":true}}'
- vLLM version: v0.10.0
- vLLM main:
9edd1db02b
---------
Signed-off-by: Wang Kunpeng <1289706727@qq.com>
Signed-off-by: SlightwindSec <slightwindsec@gmail.com>
Co-authored-by: SlightwindSec <slightwindsec@gmail.com>
chunkprefill mla only support eager mode now，we want to optimaze it by
support torchair graph, the idea is simple, when all the request is
running in decode, use torchair graph to deal with it, else when
chunkprefill or prefill only, use the eager mode
- vLLM version: v0.10.0
- vLLM main:
ebf7605b0d
Signed-off-by: haojiangzheng <justineric096@gmail.com>
Co-authored-by: haojiangzheng <justineric096@gmail.com>
### What this PR does / why we need it?
Support MTP with：
- [x] V0 Scheduler
- [x] TorchAir
- [x] Single DP
- [x] Multi DP
- [x] Disaggregate PD
Known issues：
- [ ] Not support V1 Scheduler (chunked prefill), will be supported in a
few weeks
- [ ] vllm v0.10.0 does not support metrics with `DP > 1` right now,
need to comment out the line 171-175 in file
`vllm/vllm/v1/metrics/loggers.py`
```
if (len(self.engine_indexes) > 1
and vllm_config.speculative_config is not None):
raise NotImplementedError("Prometheus metrics with Spec Decoding "
"with >1 EngineCore per AsyncLLM is not "
"supported yet.")
```
To start an online server with torchair enabled, here is an example:
```
python -m vllm.entrypoints.openai.api_server \
--model="/weights/DeepSeek-R1_w8a8/" \
--trust-remote-code \
--max-model-len 40000 \
--tensor-parallel-size 4 \
--data_parallel_size 4 \
--max-num-seqs 16 \
--no-enable-prefix-caching \
--enable_expert_parallel \
--served-model-name deepseekr1 \
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}' \
--quantization ascend \
--host 0.0.0.0 \
--port 1234 \
--additional-config '{"ascend_scheduler_config":{"enabled":true,"enable_chunked_prefill":false},"torchair_graph_config":{"enabled":true,"graph_batch_sizes":[16]},"enable_weight_nz_layout":true}' \
--gpu_memory_utilization 0.9
```
offline example with torchair enabled
```
from vllm import LLM, SamplingParams
prompts = [
"Hello, my name is",
"The president of the United States is",
"The capital of France is",
"The future of AI is",
]
# Create a sampling params object.
sampling_params = SamplingParams(max_tokens=16, temperature=0)
# Create an LLM.
llm = LLM(
model="/home/data/DeepSeek-R1_w8a8/",
tensor_parallel_size=16,
max_num_seqs=16,
gpu_memory_utilization=0.9,
distributed_executor_backend="mp",
enable_expert_parallel=True,
speculative_config={
"method": "deepseek_mtp",
"num_speculative_tokens": 1,
},
trust_remote_code=True,
enforce_eager=False,
max_model_len=2000,
additional_config = {
'torchair_graph_config': {
'enabled': True,
"graph_batch_sizes": [16],
'enable_multistream_shared_expert': False,
},
"ascend_scheduler_config": {
"enabled": True
},
# 'expert_tensor_parallel_size': 16,
}
)
# Generate texts from the prompts.
# llm.start_profile()
outputs = llm.generate(prompts, sampling_params)
# llm.stop_profile()
for output in outputs:
prompt = output.prompt
generated_text = output.outputs[0].text
print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
```
- vLLM version: v0.10.0
- vLLM main:
302962e806
---------
Signed-off-by: xuyexiong <xuyexiong@huawei.com>
### What this PR does / why we need it?
Add ut for files in folder /attention
### Does this PR introduce _any_ user-facing change?
No
- vLLM version: v0.10.0
- vLLM main:
139a7f07bd
---------
Signed-off-by: lwq <liwenquan5@huawei.com>
Co-authored-by: lwq <liwenquan5@huawei.com>
### What this PR does / why we need it?
- Upgrade to v0.10.0
- Drop v0.9.2 version compatibility
- Add patch for
`vllm_ascend/patch/worker/patch_common/patch_sampler_gather_logprobs.py`
as workaround of
f3a683b7c9
for v0.10.0 and also add e2e test `test_models_prompt_logprobs`
- Pin transformers<4.54.0 as workaround of
https://github.com/vllm-project/vllm-ascend/issues/2034
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
- Test locally:
`VLLM_USE_MODELSCOPE=true pytest -sv
tests/e2e/singlecard/test_offline_inference.py::test_models_prompt_logprobs`
- CI passed
- vLLM version: v0.9.2
- vLLM main:
7728dd77bb
---------
Signed-off-by: Yikun Jiang <yikunkero@gmail.com>
Before do attention module refactor, we can do some code cleanup to make
the next step easier.
What this PR does:
1. remove uesless `common_prefix_len` for attention builder
2. remove uesless `is_only_prefill` and `num_input_tokens` in attention
metadata.
3. remove `CommonAttentionMetadata` and ues `query_start_loc` instead,
`CommonAttentionMetadata` is over designed and uesless
4. update the attention backend input parameters to keep the same as
vLLM.
5. Rename attention name to the same style with `ASCEND` prefix
- vLLM version: v0.9.2
- vLLM main:
107111a859
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
### What this PR does / why we need it?
Support pipeline parallel with ray backend in V1Engine.
Fixes#1751
### Does this PR introduce _any_ user-facing change?
Users could specify ray as distributed backend when inferencing with pp
### How was this patch tested?
CI passed with new added test.
- vLLM version: v0.9.2
- vLLM main:
32142b3c62
---------
Signed-off-by: MengqingCao <cmq0113@163.com>
vLLM commit
752c6ade2e
removed `blocksparse_params` for attention backend. This PR does the
same change to make CI happy.
- vLLM version: v0.9.2
- vLLM main:
9499e26e2a
---------
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Signed-off-by: Yikun Jiang <yikunkero@gmail.com>
Co-authored-by: Yikun Jiang <yikunkero@gmail.com>
### What this PR does / why we need it?
If a small batch of short requests is sent first, forming a chunk with a
length <128, it will corrupt the `attn_mask_cache`, causing subsequent
requests that do not form a chunk to have accuracy issues.
The root cause of this problem is the use of in-place multiplication.
Modifying it to use out-of-place multiplication will resolve the
accuracy problem.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
Yes.
- vLLM version: v0.9.2
- vLLM main:
ad6c2e1a0b
---------
Signed-off-by: ApsarasX <apsarax@outlook.com>
### What this PR does / why we need it?
Use Base test and cleanup all manaul patch code
- Cleanup EPLB config to avoid tmp test file
- Use BaseTest with global cache
- Add license
- Add a doc to setup unit test in local env
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
CI passed
Signed-off-by: Yikun Jiang <yikunkero@gmail.com>
### What this PR does / why we need it?
1. drop some useless code for w8a8 fusedmoe
2. Add in8 kv cache check
3. Add more ut.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
CI passed with new added test.
---------
Signed-off-by: zhuyilin <809721801@qq.com>
Signed-off-by: tianyitang <tangtianyi4@huawei.com>
Co-authored-by: tianyitang <tangtianyi4@huawei.com>
