Modelpublic

skillsafe-ai/bge-base-en-v1.5

sourceHugging Facemitupdated 4d agoView on Hugging Face

0likes34downloads

Model Card

BGE base en v1.5 embeddings (768-d)

Browser-ready import artifacts for feature-extraction, produced by SkillSafe's reproducible converter (`models/` in skillsafe.ai) from a pinned upstream source. Every byte here is derivable from that source plus the recipe below; nothing was edited by hand.

Provenance


Upstream	https://huggingface.co/BAAI/bge-base-en-v1.5/tree/a5beb1e3e68b9ab74eb54cfd186867f64f240e1a
Upstream SHA-256 / commit	`a5beb1e3e68b9ab74eb54cfd186867f64f240e1a`
Recipe	`recipes/bge-base-en-v1.5.yaml` — sha256 `e18fa3565b0987daf57fd42046ed70c0039a7e7d5fd5b9f311f624f4ea42dda8`
Toolchain	Python 3.12.13, torch 2.10.0, onnx 1.23.0, onnxruntime 1.30.0 on Darwin 25.6.0 arm64
Converted	2026-09-22T20:35:51+00:00

Files

file	class	size	SHA-256
`1_Pooling/config.json`	bundle	0.00 MB	`c9bef85e8bbf4b2eab4941b3fb62bd33f88686748b478f2e264d256472d9643b`
`config.json`	bundle	0.00 MB	`bc00af31a4a31b74040d73370aa83b62da34c90b75eb77bfa7db039d90abd591`
`onnx/model.onnx`	registry	415.62 MB	`9bc579acdba21c253c62a9bf866891355a63ffa3442b52c8a37d75b2ccb91848`
`special_tokens_map.json`	bundle	0.00 MB	`b6d346be366a7d1d48332dbc9fdf3bf8960b5d879522b7799ddba59e76237ee3`
`tokenizer.json`	bundle	0.68 MB	`d241a60d5e8f04cc1b2b3e9ef7a4921b27bf526d9f6050ab90f9267a1f9e5c66`
`tokenizer_config.json`	bundle	0.00 MB	`9261e7d79b44c8195c1cada2b453e55b00aeb81e907a6664974b4d7776172ab3`

registry files are parameter files served from models.skillsafe.ai once vetted; bundle files ship inside an app; registry-shared is a runtime library reused by every model of the same architecture.

Verification

Imported as published upstream (no conversion). Each file is pinned by SHA-256 to its source; every ONNX file passed onnx.checker and a CPU smoke run under onnxruntime with zero-filled inputs at the declared shapes:

file	inputs	outputs	ms
`onnx/model.onnx`	inputids[1, 8], attentionmask[1, 8], tokentypeids[1, 8]	lasthiddenstate[1, 8, 768]	6.0

Use in the browser

import * as ort from "onnxruntime-web";
const session = await ort.InferenceSession.create("https://huggingface.co/skillsafe-ai/bge-base-en-v1.5/resolve/main/onnx/model.onnx", { executionProviders: ["webgpu", "wasm"] });

Contract (onnx/model.onnx): input input_ids int64 ['batch_size', 'sequence_length'], attention_mask int64 ['batch_size', 'sequence_length'], token_type_ids int64 ['batch_size', 'sequence_length'] → output last_hidden_state float32 ['batch_size', 'sequence_length', 768]. Opset 11.

Licence and attribution

bge-base-en-v1.5: Beijing Academy of Artificial Intelligence (FlagEmbedding), MIT License. https://huggingface.co/BAAI/bge-base-en-v1.5 — the repo's own ONNX export.

Licence: MIT — notice: https://huggingface.co/BAAI/bge-base-en-v1.5/blob/a5beb1e3e68b9ab74eb54cfd186867f64f240e1a/README.md. The conversion recipe and this model card are part of the SkillSafe repository and carry its licence; the weights remain under the upstream licence above.

The full manifest.json in this repo records the recipe, sources, toolchain (including the uv.lock hash) and per-file verification numbers.