From 79f096d43bff852ea7b8e186b7504bd2868d012b Mon Sep 17 00:00:00 2001
From: loading66 <102725781+loading66@users.noreply.github.com>
Date: Sat, 11 Jul 2026 15:21:29 +0800
Subject: [PATCH] [DOCS][NPU]update npu support features and models (#30843)
---
.../ascend_npu_support_features.mdx | 7 ++--
.../ascend-npus/ascend_npu_support_models.mdx | 39 -------------------
2 files changed, 4 insertions(+), 42 deletions(-)
diff --git a/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_features.mdx b/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_features.mdx
index d022fe465..82e5a6abd 100644
--- a/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_features.mdx
+++ b/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_features.mdx
@@ -1247,7 +1247,7 @@ click [Server Arguments](../../advanced_features/server_arguments).
`--lora-strict-loading` |
`False` |
Type: bool |
- A2, A3 |
+ Special for GPU |
| `--lora-drain-wait-threshold` |
@@ -1638,7 +1638,8 @@ click [Server Arguments](../../advanced_features/server_arguments).
| `--deepep-dispatcher-output-dtype` |
`auto` |
- `auto`, `bf16`, `int8` |
+ `auto`, `bf16`, `int8` (When enabling DeepEP for a quantized model, set the dispatcher output dtype according to your model’s quantization.
+If the value is int8, you must also set the environment variable:DEEP_NORMAL_MODE_USE_INT8_QUANT=1) |
A2, A3 |
@@ -2894,7 +2895,7 @@ click [Server Arguments](../../advanced_features/server_arguments).
| `--enable-quant-communications` |
`False` |
- bool flag (set to enable) |
+ bool flag (set to enable) (When using DeepEP with a quantized model, you must set --deepep-dispatcher-output-dtype) |
A2, A3 |
diff --git a/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_models.mdx b/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_models.mdx
index 71c3455ae..e372977f2 100644
--- a/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_models.mdx
+++ b/docs_new/docs/hardware-platforms/ascend-npus/ascend_npu_support_models.mdx
@@ -571,45 +571,6 @@ You are welcome to enable various models based on your business requirements.
-## Diffusion language models
-
-
-
-
-
-
-
-
-
-
- | Models |
- Model Family |
- A2 Supported |
- A3 Supported |
-
-
-
-
- | inclusionAI/LLaDA2.0-flash |
- LLaDA2.0 (mini, flash) |
- ✅ |
- ✅ |
-
-
- | JetLM/SDAR-8B-Chat |
- SDAR (JetLM) |
- ✅ |
- ✅ |
-
-
- | JetLM/SDAR-30B-A3B-Chat |
- SDAR (JetLM) |
- ✅ |
- ✅ |
-
-
-
-
## Embedding Models