Skip to content

Commit 79b2c6d

Browse files
authored
Merge pull request #11 from BaizeAI/feat/qwen3-235b-a22b
feat: add qwen3-235b-a22b
2 parents 5ac34f4 + 74eeb54 commit 79b2c6d

1 file changed

Lines changed: 38 additions & 0 deletions

File tree

Lines changed: 38 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,38 @@
1+
apiVersion: model.hydra.io/v1alpha1
2+
kind: ModelSpec
3+
metadata:
4+
name: qwen3-235b-a22b-fp8
5+
spec:
6+
deployments:
7+
- customRuntimeArgs: []
8+
resourceRequirements:
9+
cpu: 8
10+
gpuCount: 8
11+
gpuType: nvidia-vgpu
12+
memory: 640
13+
perGPUMemoryGB: 80
14+
runtime: vllm
15+
versionRequired: '>=0.8.5'
16+
descriptor:
17+
description:
18+
enUS: Qwen3 is the latest generation of large language models in Qwen series, offering a comprehensive suite of dense and mixture-of-experts (MoE) models. Built upon extensive training, Qwen3 delivers groundbreaking advancements in reasoning, instruction-following, agent capabilities, and multilingual support.
19+
zhCN: Qwen3是通义系列最新一代大语言模型,提供了一整套稠密和混合专家(MoE)模型。基于大量训练,Qwen3在推理、指令跟随、智能体能力和多语言支持方面取得了突破性进展。
20+
display: Qwen3-235B-A22B-FP8
21+
icon:
22+
src: https://public-resources.d.run/models/logos/qwen-model-logo.svg
23+
type: image/svg
24+
links:
25+
- description: About
26+
url: https://github.com/QwenLM
27+
provider:
28+
id: alibaba
29+
name:
30+
enUS: Alibaba
31+
zhCN: 通义千问
32+
tags:
33+
- TEXT_GENERATION
34+
source:
35+
huggingface:
36+
name: Qwen/Qwen3-235B-A22B-FP8
37+
modelscope:
38+
name: Qwen/Qwen3-235B-A22B-FP8

0 commit comments

Comments
 (0)