|
|
ffb4de7642
|
命令如下
docker run -dit --name gpt-120b \
-e NVIDIA_VISIBLE_DEVICES=0,1 \
--shm-size 64g \
--network=host -v ./:/data/models \
--restart=always \
lmsysorg/sglang:dev python3 -m sglang.launch_server --model /data/models --host 0.0.0.0 --port 40002 --trust-remote --tp 2 --served-model-name gpt-120b --tool-call-parser gpt-oss
自己替换下模型ID，针对模型部署和评测详细感兴趣的可以微信搜索 AI云枢 有详细讲解和其他模型的部署评测教程
|
2025-08-15 01:14:46 +00:00 |
|
|
|
6d460b1f00
|
Configure <|call|> as an EOS token (#109)
- Configure `<|call|>` as an EOS token (a44a7f14988426860a0a8145cf2bdfda2d7cadac)
Co-authored-by: Pedro Cuenca <pcuenq@users.noreply.huggingface.co>
|
2025-08-14 11:47:38 +00:00 |
|
|
|
110e39cb08
|
Update README.md
Clarify MXFP4 quantization callout
|
2025-08-13 13:12:36 +00:00 |
|
|
|
00330d7e17
|
Update chat_template.jinja (#94)
- Update chat_template.jinja (6965815ae896176f6b820fd508d15ebbca89a903)
Co-authored-by: Quentin Gallouédec <qgallouedec@users.noreply.huggingface.co>
|
2025-08-09 08:16:19 +08:00 |
|
|
|
c5fe96c09a
|
Updated chat template (#82)
- Updated chat template (03fd454aae9be5799e9531726db7f6d0673675cb)
Co-authored-by: Matthew Carrigan <Rocketknight1@users.noreply.huggingface.co>
|
2025-08-08 06:14:10 +08:00 |
|
|
|
d62b81784f
|
Update the model name in USAGE_POLICY (#52)
- Update USAGE_POLICY (ccef4ca7b48a5797b0b436e77f5cd0c643449942)
Co-authored-by: B. <Enes@users.noreply.huggingface.co>
|
2025-08-07 06:46:22 +08:00 |
|
|
|
bfbcca03ca
|
Upload folder using ModelScope SDK
|
2025-08-05 18:38:51 +00:00 |
|
|
|
4a757648f3
|
Upload folder using ModelScope SDK
|
2025-08-05 18:38:50 +00:00 |
|
|
|
0422a104e2
|
System update meta information
|
2025-08-05 18:30:22 +00:00 |
|
|
|
b435140cbb
|
System init .gitattributes
|
2025-08-05 18:30:21 +00:00 |
|
