A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 22:15 UTC
[Doc] Add clarification note regarding paligemma (#13511)
Roger Wang committed
Feb 18, 2025 at 22:24 UTC
fd84857f646e5591c16c1e650a6fe68931efd337
1 file changed
+5
-1
docs/source/models/supported_models.md
+5
-1
@@ -808,7 +808,7 @@ See [this page](#generative-models) for more information on how to use generativ
808
*
809
* ✅︎
810
* ✅︎
811
-- * `PaliGemmaForConditionalGeneration`
811
+- * `PaliGemmaForConditionalGeneration`\*
812
* PaliGemma, PaliGemma 2
813
* T + I<sup>E</sup>
814
* `google/paligemma-3b-pt-224`, `google/paligemma-3b-mix-224`, `google/paligemma2-3b-ft-docci-448`, etc.
@@ -885,6 +885,10 @@ The official `openbmb/MiniCPM-V-2` doesn't work yet, so we need to use a fork (`
885
For more details, please see: <gh-pr:4087#issuecomment-2250397630>
886
:::
887
888
+:::{note}
889
+Currently the PaliGemma model series is implemented without PrefixLM attention mask. This model series may be deprecated in a future release.
890
+:::
891
+
892
:::{note}
893
`mistral-community/pixtral-12b` does not support V1 yet.
894
:::