You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/source/faqs.md
+20-1Lines changed: 20 additions & 1 deletion
Original file line number
Diff line number
Diff line change
@@ -55,7 +55,7 @@ After configuration, you can download our container from `m.daocloud.io/quay.io/
55
55
56
56
### 3. What models does vllm-ascend supports?
57
57
58
-
Currently, we have already fully tested and supported `Qwen` / `Deepseek` (V0 only) / `Llama` models, other models we have tested are shown [<u>here</u>](https://vllm-ascend.readthedocs.io/en/latest/user_guide/supported_models.html). Plus, accoding to users' feedback, `gemma3` and `glm4` are not supported yet. Besides, more models need test.
58
+
Currently, we have already fully tested and supported `Qwen` / `Deepseek` (V0 only) / `Llama` models, other models we have tested are shown [<u>here</u>](https://vllm-ascend.readthedocs.io/en/latest/user_guide/supported_models.html). Plus, according to users' feedback, `gemma3` and `glm4` are not supported yet. Besides, more models need test.
59
59
60
60
### 4. How to get in touch with our community?
61
61
@@ -69,3 +69,22 @@ There are many channels that you can communicate with our community developers /
69
69
### 5. What features does vllm-ascend V1 supports?
70
70
71
71
Find more details [<u>here</u>](https://github.com/vllm-project/vllm-ascend/issues/414).
72
+
73
+
### 6. How to solve the problem of "Failed to infer device type" or "libatb.so: cannot open shared object file"?
74
+
75
+
Basicly, the reason is that the NNAL environment is not sourced. Please try `source /usr/local/Ascend/nnal/atb/set_env.sh` to solve the problem.
76
+
77
+
### 7. Does vllm-ascend support Atlas 300I Duo?
78
+
79
+
No, vllm-ascend now only supports Atlas A2 series. We are working on it.
80
+
81
+
### 8. How does vllm-ascend perform?
82
+
83
+
Currently, only some models are imporved. Such as `Qwen2 VL`, `Deepseek V3`. Others are not good enough. In the future, we will support graph mode and custom ops to improve the performance of vllm-ascend. And when the official release of vllm-ascend is released, you can install `mindie-turbo` with `vllm-ascend` to speed up the inference as well.
84
+
85
+
### 9. How vllm-ascend work with vllm?
86
+
vllm-ascend is a plugin for vllm. Basicly, the version of vllm-ascend is the same as the version of vllm. For example, if you use vllm 0.7.3, you should use vllm-ascend 0.7.3 as well. For main branch, we will make sure `vllm-ascend` and `vllm` are compatible by each commit.
87
+
88
+
### 10. Does vllm-ascend support Prefill Disaggregation feature?
89
+
90
+
Currently, only 1P1D is supported by vllm. For vllm-ascend, it'll be done by [this PR](https://github.com/vllm-project/vllm-ascend/pull/432). For NPND, vllm is not stable and fully supported yet. We will make it stable and supported by vllm-ascend in the future.
0 commit comments