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 14, 05:30 UTC
[docs] Prompt Embedding feature support (#25288)
Signed-off-by: Andrew Sansom <andrew@protopia.ai>
Andrew Sansom committed
Sep 19, 2025 at 19:46 UTC
b8a287a0a8a035073d67b5101687e3a753dd02ac
2 files changed
+18
-19
docs/features/README.md
+18
-16
@@ -36,22 +36,23 @@ th:not(:first-child) {
36
}
37
</style>
38
39
-| Feature | [CP][chunked-prefill] | [APC](automatic_prefix_caching.md) | [LoRA](lora.md) | [SD](spec_decode.md) | CUDA graph | [pooling](../models/pooling_models.md) | <abbr title="Encoder-Decoder Models">enc-dec</abbr> | <abbr title="Logprobs">logP</abbr> | <abbr title="Prompt Logprobs">prmpt logP</abbr> | <abbr title="Async Output Processing">async output</abbr> | multi-step | <abbr title="Multimodal Inputs">mm</abbr> | best-of | beam-search |
40
-|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
41
-| [CP][chunked-prefill] | ✅ | | | | | | | | | | | | | |
42
-| [APC](automatic_prefix_caching.md) | ✅ | ✅ | | | | | | | | | | | | |
43
-| [LoRA](lora.md) | ✅ | ✅ | ✅ | | | | | | | | | | | |
44
-| [SD](spec_decode.md) | ✅ | ✅ | ❌ | ✅ | | | | | | | | | | |
45
-| CUDA graph | ✅ | ✅ | ✅ | ✅ | ✅ | | | | | | | | | |
46
-| [pooling](../models/pooling_models.md) | 🟠\* | 🟠\* | ✅ | ❌ | ✅ | ✅ | | | | | | | | |
47
-| <abbr title="Encoder-Decoder Models">enc-dec</abbr> | ❌ | [❌](gh-issue:7366) | ❌ | [❌](gh-issue:7366) | ✅ | ✅ | ✅ | | | | | | | |
48
-| <abbr title="Logprobs">logP</abbr> | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | | | | | | |
49
-| <abbr title="Prompt Logprobs">prmpt logP</abbr> | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | ✅ | | | | | |
50
-| <abbr title="Async Output Processing">async output</abbr> | ✅ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ | | | | |
51
-| multi-step | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ | ✅ | | | |
52
-| [mm](multimodal_inputs.md) | ✅ | ✅ | [🟠](gh-pr:4194)<sup>^</sup> | ❔ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❔ | ✅ | | |
53
-| best-of | ✅ | ✅ | ✅ | [❌](gh-issue:6137) | ✅ | ❌ | ✅ | ✅ | ✅ | ❔ | [❌](gh-issue:7968) | ✅ | ✅ | |
54
-| beam-search | ✅ | ✅ | ✅ | [❌](gh-issue:6137) | ✅ | ❌ | ✅ | ✅ | ✅ | ❔ | [❌](gh-issue:7968) | ❔ | ✅ | ✅ |
39
+| Feature | [CP][chunked-prefill] | [APC](automatic_prefix_caching.md) | [LoRA](lora.md) | [SD](spec_decode.md) | CUDA graph | [pooling](../models/pooling_models.md) | <abbr title="Encoder-Decoder Models">enc-dec</abbr> | <abbr title="Logprobs">logP</abbr> | <abbr title="Prompt Logprobs">prmpt logP</abbr> | <abbr title="Async Output Processing">async output</abbr> | multi-step | <abbr title="Multimodal Inputs">mm</abbr> | best-of | beam-search | [prompt-embeds](prompt_embeds.md) |
40
+|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
41
+| [CP][chunked-prefill] | ✅ | | | | | | | | | | | | | | |
42
+| [APC](automatic_prefix_caching.md) | ✅ | ✅ | | | | | | | | | | | | | |
43
+| [LoRA](lora.md) | ✅ | ✅ | ✅ | | | | | | | | | | | | |
44
+| [SD](spec_decode.md) | ✅ | ✅ | ❌ | ✅ | | | | | | | | | | | |
45
+| CUDA graph | ✅ | ✅ | ✅ | ✅ | ✅ | | | | | | | | | | |
46
+| [pooling](../models/pooling_models.md) | 🟠\* | 🟠\* | ✅ | ❌ | ✅ | ✅ | | | | | | | | | |
47
+| <abbr title="Encoder-Decoder Models">enc-dec</abbr> | ❌ | [❌](gh-issue:7366) | ❌ | [❌](gh-issue:7366) | ✅ | ✅ | ✅ | | | | | | | | |
48
+| <abbr title="Logprobs">logP</abbr> | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | | | | | | | |
49
+| <abbr title="Prompt Logprobs">prmpt logP</abbr> | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ✅ | ✅ | ✅ | | | | | | |
50
+| <abbr title="Async Output Processing">async output</abbr> | ✅ | ✅ | ✅ | ❌ | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ | | | | | |
51
+| multi-step | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ❌ | ✅ | ✅ | ✅ | ✅ | | | | |
52
+| [mm](multimodal_inputs.md) | ✅ | ✅ | [🟠](gh-pr:4194)<sup>^</sup> | ❔ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❔ | ✅ | | | |
53
+| best-of | ✅ | ✅ | ✅ | [❌](gh-issue:6137) | ✅ | ❌ | ✅ | ✅ | ✅ | ❔ | [❌](gh-issue:7968) | ✅ | ✅ | | |
54
+| beam-search | ✅ | ✅ | ✅ | [❌](gh-issue:6137) | ✅ | ❌ | ✅ | ✅ | ✅ | ❔ | [❌](gh-issue:7968) | ❔ | ✅ | ✅ | |
55
+| [prompt-embeds](prompt_embeds.md) | ✅ | [❌](gh-issue:25096) | ? | ❌ | ✅ | ❌ | ❌ | ✅ | ❌ | ? | ? | ❌ | ? | ? | ✅ |
56
57
\* Chunked prefill and prefix caching are only applicable to last-token pooling.
58
<sup>^</sup> LoRA is only applicable to the language backbone of multimodal models.
@@ -76,3 +77,4 @@ th:not(:first-child) {
77
| multi-step | ✅ | ✅ | ✅ | ✅ | ✅ | [❌](gh-issue:8477) | ✅ | ❌ |
78
| best-of | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
79
| beam-search | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
80
+| [prompt-embeds](prompt_embeds.md) | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ? | [❌](gh-issue:25097) |
docs/features/prompt_embeds.md
-3
@@ -6,9 +6,6 @@ This page teaches you how to pass prompt embedding inputs to vLLM.
6
7
The traditional flow of text data for a Large Language Model goes from text to token ids (via a tokenizer) then from token ids to prompt embeddings. For a traditional decoder-only model (such as meta-llama/Llama-3.1-8B-Instruct), this step of converting token ids to prompt embeddings happens via a look-up from a learned embedding matrix, but the model is not limited to processing only the embeddings corresponding to its token vocabulary.
8
9
-!!! note
10
- Prompt embeddings are currently only supported in the v0 engine.
11
-
9
## Offline Inference
10
11
To input multi-modal data, follow this schema in [vllm.inputs.EmbedsPrompt][]: