iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Model] Add zerank-2 reranker (#60415)

Browse Source 
Signed-off-by: rdlhio <contact@rdlh.io>
This commit is contained in:
Rémi Delhaye
2026-10-09 03:01:51 +00:00
committed by GitHub
parent 702ce313f4
commit a1cd0ec650
5 changed files with 119 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/models/pooling_models/scoring.md
+8 -1
View File
@@ -49,7 +49,7 @@ The three supported scoring functions are as illustrated in the figure below.
| `LlamaBidirectionalForSequenceClassification`<sup>C</sup> | Llama-based with bidirectional attention | `nvidia/llama-nemotron-rerank-1b-v2`, etc. | [nemotron-rerank.jinja](../../../examples/pooling/score/template/nemotron-rerank.jinja) | ✅︎ | ✅︎ |
| `ModernBertForSequenceClassification` | ModernBERT-based | `Alibaba-NLP/gte-reranker-modernbert-base`, etc. | N/A | | |
| `Qwen2ForSequenceClassification`<sup>C</sup> | Qwen2-based | `mixedbread-ai/mxbai-rerank-base-v2`(see note), etc. | [mxbai_rerank_v2.jinja](../../../examples/pooling/score/template/mxbai_rerank_v2.jinja) | ✅︎ | ✅︎ |
| `Qwen3ForSequenceClassification`<sup>C</sup> | Qwen3-based | `tomaarsen/Qwen3-Reranker-0.6B-seq-cls`, `Qwen/Qwen3-Reranker-0.6B`(see note), etc. | [qwen3_reranker.jinja](../../../examples/pooling/score/template/qwen3_reranker.jinja) | ✅︎ | ✅︎ |
| `Qwen3ForSequenceClassification`<sup>C</sup> | Qwen3-based | `tomaarsen/Qwen3-Reranker-0.6B-seq-cls`, `Qwen/Qwen3-Reranker-0.6B`(see note), `polaria-tech/zerank-2-reranker-vllm`(see note), etc. | [qwen3_reranker.jinja](../../../examples/pooling/score/template/qwen3_reranker.jinja), [zerank2.jinja](../../../examples/pooling/score/template/zerank2.jinja) | ✅︎ | ✅︎ |
| `RobertaForSequenceClassification` | RoBERTa-based | `cross-encoder/quora-roberta-base`, etc. | N/A | ✅︎ | |
| `XLMRobertaForSequenceClassification` | XLM-RoBERTa-based | `BAAI/bge-reranker-v2-m3`, etc. | N/A | ✅︎ | |
| `*Model`<sup>C</sup>, `*ForCausalLM`<sup>C</sup>, etc. | Generative models | N/A | N/A | \* | \* |
@@ -88,6 +88,13 @@ The three supported scoring functions are as illustrated in the figure below.
vllm serve Qwen/Qwen3-Reranker-0.6B --hf_overrides '{"architectures": ["Qwen3ForSequenceClassification"],"classifier_from_token": ["no", "yes"],"is_original_qwen3_reranker": true}'
```
!!! note
`polaria-tech/zerank-2-reranker-vllm` is [zeroentropy/zerank-2-reranker](https://huggingface.co/zeroentropy/zerank-2-reranker) converted to `Qwen3ForSequenceClassification`; it returns zerank-2's calibrated score `sigmoid(logit / 5)`. Serve it with [zerank2.jinja](../../../examples/pooling/score/template/zerank2.jinja), which also supports the `instruction` request field.
```bash
vllm serve polaria-tech/zerank-2-reranker-vllm --chat-template examples/pooling/score/template/zerank2.jinja
```
#### Multimodal Models
!!! note