gliner2-multi-v1-agent-ONNX

GLiNER2 multi v1, one-graph ONNX for the browser#

ONNX export of fastino/gliner2-multi-v1 (mDeBERTa-v3-base, multilingual) as a single graph that serves both entity extraction and label classification, for Transformers.js on WebGPU. It was made for Zipline, a browser agent that runs entirely in Chrome, and covers the two calls such an agent makes: pulling values out of a goal (extract_entities) and scoring page controls against it (classify with a softmax over labels).

Inputs and outputs#

Entity decoding is the library's default: sigmoid ≥ 0.5, map word spans to characters, then keep spans greedily by confidence without character overlap. Entity extraction only reads instance 0, so the count GRU is unrolled once (export_onnx.py asserts that instance 0 does not depend on the unroll length).

Verification#

conversion/reference.json holds 14 calls recorded from the Python gliner2 library (six goals across English and French, eight control-scoring calls on flight-search and maps pages), with their exact token ids. Against it:

The JavaScript runtime reproduces the token ids exactly. On WebGPU (fp16, Apple M3 Pro): 35–37 ms per extraction with nine entity types, 43–44 ms per classification with twelve labels.

Rebuilding#

pip install torch "gliner2[local]" onnx onnxruntime onnxscript transformers
python conversion/reference.py        # Python reference outputs
python conversion/export_onnx.py      # fp32 graph
python conversion/convert_fp16.py     # fp16 graph
python conversion/verify_onnx.py      # parity check

(The scripts write to dist-model/; external data is renamed to the .onnx_data convention afterwards.)

Credits#

Model by Fastino (GLiNER2), Apache-2.0. The agent calls these outputs serve come from gliner2-ultrafast (MIT).