EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

remove useless patch (#4699)

Browse Source 
patach_config is useless now. Let's remove it


- vLLM version: v0.12.0
- vLLM main:
ad32e3e19c

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Co-authored-by: Mengqing Cao <cmq0113@163.com>
This commit is contained in:
wangxiyuan
2025-12-08 11:02:42 +08:00
committed by GitHub
parent 866347a621
commit 0b65ac6c4b
21 changed files with 30 additions and 277 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
tests/e2e/nightly/features/test_mtpx_deepseek_r1_0528_w8a8.py
View File

@@ -74,10 +74,7 @@ async def test_models(model: str, mode: str) -> None:
"VLLM_EXECUTE_MODEL_TIMEOUT_SECONDS": "3600000"
}
additional_config: dict[str, Any] = {}
speculative_config = {
"num_speculative_tokens": 2,
"method": "deepseek_mtp"
}
speculative_config = {"num_speculative_tokens": 2, "method": "mtp"}
compilation_config = {
"cudagraph_capture_sizes": [56],
"cudagraph_mode": "FULL_DECODE_ONLY"

5
tests/e2e/nightly/features/test_prefix_cache_deepseek_r1_0528_w8a8.py
View File

@@ -84,10 +84,7 @@ async def test_models(model: str) -> None:
"chunked_prefill_for_mla": True,
"enable_weight_nz_layout": True
}
speculative_config = {
"num_speculative_tokens": 1,
"method": "deepseek_mtp"
}
speculative_config = {"num_speculative_tokens": 1, "method": "mtp"}
server_args = [
"--quantization", "ascend", "--data-parallel-size", "2",
"--tensor-parallel-size", "8", "--enable-expert-parallel", "--port",

5
tests/e2e/nightly/models/test_deepseek_r1_0528_w8a8.py
View File

@@ -76,10 +76,7 @@ async def test_models(model: str, mode: str) -> None:
"HCCL_BUFFSIZE": "1024",
"PYTORCH_NPU_ALLOC_CONF": "expandable_segments:True"
}
speculative_config = {
"num_speculative_tokens": 1,
"method": "deepseek_mtp"
}
speculative_config = {"num_speculative_tokens": 1, "method": "mtp"}
additional_config = {
"torchair_graph_config": {
"enabled": True,

5
tests/e2e/nightly/models/test_deepseek_r1_w8a8_eplb.py
View File

@@ -62,10 +62,7 @@ async def test_models(model: str) -> None:
"DISABLE_L2_CACHE": "1",
"DYNAMIC_EPLB": "true",
}
speculative_config = {
"num_speculative_tokens": 1,
"method": "deepseek_mtp"
}
speculative_config = {"num_speculative_tokens": 1, "method": "mtp"}
compilation_config = {
"cudagraph_capture_sizes": [24],
"cudagraph_mode": "FULL_DECODE_ONLY"

4
tests/e2e/nightly/multi_node/config/models/DeepSeek-R1-W8A8-A2-torchair.yaml
View File

@@ -29,7 +29,7 @@ deployment:
--trust-remote-code
--quantization ascend
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--additional-config '{"torchair_graph_config":{"enabled":true,"enable_multistream_moe":true},"chunked_prefill_for_mla":true,"enable_weight_nz_layout":true}'
-
@@ -50,7 +50,7 @@ deployment:
--trust-remote-code
--quantization ascend
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--additional-config '{"torchair_graph_config":{"enabled":true,"enable_multistream_moe":true},"chunked_prefill_for_mla":true,"enable_weight_nz_layout":true}'
benchmarks:
acc:

4
tests/e2e/nightly/multi_node/config/models/DeepSeek-R1-W8A8-A2.yaml
View File

@@ -30,7 +30,7 @@ deployment:
--quantization ascend
--gpu-memory-utilization 0.9
--enforce-eager
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--additional-config '{"chunked_prefill_for_mla":true,"enable_weight_nz_layout":true}'
-
@@ -52,6 +52,6 @@ deployment:
--quantization ascend
--gpu-memory-utilization 0.9
--enforce-eager
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--additional-config '{"chunked_prefill_for_mla":true,"enable_weight_nz_layout":true}'
benchmarks:

8
tests/e2e/nightly/multi_node/config/models/DeepSeek-R1-W8A8-EPLB.yaml
View File

@@ -39,7 +39,7 @@ deployment:
--max-num-batched-tokens 16384
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -69,7 +69,7 @@ deployment:
--max-num-batched-tokens 16384
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -100,7 +100,7 @@ deployment:
--max-num-batched-tokens 256
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -130,7 +130,7 @@ deployment:
--max-num-batched-tokens 256
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",

8
tests/e2e/nightly/multi_node/config/models/DeepSeek-R1-W8A8.yaml
View File

@@ -38,7 +38,7 @@ deployment:
--max-num-batched-tokens 16384
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -68,7 +68,7 @@ deployment:
--max-num-batched-tokens 16384
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -99,7 +99,7 @@ deployment:
--max-num-batched-tokens 256
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",
@@ -129,7 +129,7 @@ deployment:
--max-num-batched-tokens 256
--trust-remote-code
--gpu-memory-utilization 0.9
--speculative-config '{"num_speculative_tokens": 1, "method":"deepseek_mtp"}'
--speculative-config '{"num_speculative_tokens": 1, "method":"mtp"}'
--kv-transfer-config
'{"kv_connector": "LLMDataDistCMgrConnector",
"kv_buffer_device": "npu",

2
tests/e2e/singlecard/spec_decode_v1/test_v1_mtp_correctness.py
View File

@@ -56,7 +56,7 @@ def mtp_correctness(sampling_config: SamplingParams,
enable_expert_parallel=True,
speculative_config={
"method":
"deepseek_mtp",
"mtp",
"num_speculative_tokens":
num_speculative_tokens,
"disable_padded_drafter_batch":

2
tests/e2e/singlecard/spec_decode_v1/test_v1_mtp_torchair_correctness.py
View File

@@ -58,7 +58,7 @@ def mtp_torchair_correctness(
distributed_executor_backend="mp",
enable_expert_parallel=True,
speculative_config={
"method": "deepseek_mtp",
"method": "mtp",
"num_speculative_tokens": 1,
},
enforce_eager=False,