A model is still running

A model is still running or downloading. Leaving will stop the current operation. Stay on this page to let it finish.

Skip to content
LocalAI.toolsBeta

Join the waitlist

Get an email when the new apps and model runs launch.

No account needed.

BGE 1.5

Compare meaning, explore text vectors, and build semantic search on your device.

Browser inferencePowered by WebAI · Private · Unlimited runs
About this model

Turn English text into 384-number vectors for meaning-based search and sentence comparison. Supply the appropriate query prefix when searching; output is vectors, not generated text.

License
MIT
Model released
Unknown
Source repository created
Sep 12, 2023
Model weights · Q4F16
36.2 MB

Best for

  • English semantic retrieval prototypes.
  • Text similarity and clustering.
  • Browser embedding benchmark baseline.

Keep in mind

  • Designed for English; cross-language retrieval is not certified.
  • Inputs beyond 512 tokens are truncated; chunk longer documents.
  • Use the upstream retrieval instruction on queries, not passages; the worker does not add prefixes automatically.
  • Cosine similarities are not calibrated probabilities or evidence that claims are true.
  • Domain-specific text and close rankings require application evaluation.
  • Embeddings may retain training-data biases; do not use similarity alone for consequential decisions.

Processed on your deviceInference runs in this browser. Input/output sharing requires consent for each run. Basic run logs include device information and approximate connection location. Model and runtime files download from external hosts. Saved downloads may be cleared by your browser.

Sources & technical details
Original model
BAAI/bge-small-en-v1.5
Release verification
Unknown
Release evidence
Exact original checkpoint public release date has not been verified. Repository creation, training, paper, artifact conversion and update dates are not release evidence.
Source updated
Feb 22, 2024
Browser model files
Xenova/bge-small-en-v1.5
Files updated
Jul 22, 2025
Artifact revision
ea104dacec62c0de699686887e3f920caeb4f3e3

Input

TEXT → VECTORS

Sample result

BGE Small English v1.5

Output

Explore what your text means

Generate vectors from one or more texts. Add multiple texts to compare their meaning.

Benchmark evidence

Saved worker evidence is separate from benchmarks run on your device. It does not guarantee speed or quality on every browser.

BGE Small English v1.5

16 configurations tested
Passed14
Failed2
TestedSep 15, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

Finite normalized distinct 384d vectors; relevant passage cosine exceeds unrelated passage by >=0.1.

PrecisionDeviceResult
Q4F16GPU · WebGPUAcceleratedPassedVery fast
Q4GPU · WebGPUAcceleratedPassedVery fast
FP16GPU · WebGPUAcceleratedPassedVery fast
FP32GPU · WebGPUAcceleratedPassedVery fast
Q8GPU · WebGPUAcceleratedPassedFast
UINT8GPU · WebGPUAcceleratedPassedFast
BNB4GPU · WebGPUAcceleratedPassedFast
INT8GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedVery fast
UINT8CPU · WASMPassedVery fast
Q8CPU · WASMPassedVery fast
INT8CPU · WASMPassedVery fast
BNB4CPU · WASMPassedVery fast
Q4CPU · WASMPassedVery fast
FP16CPU · WASMFailed—
Q4F16CPU · WASMFailed—
Test setup & sample details
Quality gate
Finite normalized distinct 384d vectors; relevant passage cosine exceeds unrelated passage by >=0.1.
Tested
2026-09-15T20:15:36.758Z
Environment
Isolated headless Chrome 151 on macOS; Transformers.js 4.2.0
Runtime
Transformers.js 4.2.0
Artifact revision
ea104dacec62c0de699686887e3f920caeb4f3e3
Fixture
{
  "file": "bge-small-en-v1.5.json",
  "url": "https://f005.backblazeb2.com/file/hubters-webai-public/assets/benchmarks/bge-small-en-v1.5.json",
  "size": 606,
  "sha1": "2f3cc9dccba6004ab0061011297a862c5bda1634",
  "source": "WebAI-authored English/Chinese retrieval examples with upstream BGE query instruction",
  "license": "MIT",
  "preprocessing": "Caller supplies query instruction; tokenizer truncation at 512 tokens; normalized CLS pooling."
}
Rerun when
artifact revision change; runtime version change; fixture or quality-gate change; browser/backend behavior change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7934246311375007,
      "negative": 0.43608810141044896,
      "margin": 0.35733652972705177,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 20520.30000001192,
      "inferenceMs": 72.09999999403954,
      "totalMs": 21847.09999999404,
      "grade": "excellent",
      "embeddingsPerSecond": 41.608876563772654
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7934244779234982,
      "negative": 0.43608809101734064,
      "margin": 0.3573363869061576,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 19949.69999998808,
      "inferenceMs": 69.5,
      "totalMs": 21249.69999998808,
      "grade": "excellent",
      "embeddingsPerSecond": 43.165467625899275
    }
  },
  {
    "precision": "fp16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 15786.300000011921,
      "grade": "failed"
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/graph/graph_utils.cc:30 int onnxruntime::graph_utils::GetIndexFromName(const Node &, const std::string &, bool) itr != node_args.end() was false. Attempting to get index by a name which does not exist:InsertedPrecisionFreeCast_/encoder/layer.11/output/LayerNorm/Constant_output_0for node: /embeddings/LayerNorm/Mul/SimplifiedLayerNormFusion/\n"
  },
  {
    "precision": "fp16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.598378441238317,
      "negative": 0.44285379270737396,
      "margin": 0.155524648530943,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 12284.800000011921,
      "inferenceMs": 68.09999999403954,
      "totalMs": 13708.40000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 44.05286343997908
    }
  },
  {
    "precision": "int8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7895325092454644,
      "negative": 0.43330974144109247,
      "margin": 0.3562227678043719,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9452.09999999404,
      "inferenceMs": 81.30000001192093,
      "totalMs": 10858.699999988079,
      "grade": "excellent",
      "embeddingsPerSecond": 36.900368998279376
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7905615544798418,
      "negative": 0.4338808018805735,
      "margin": 0.3566807525992683,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9614.800000011921,
      "inferenceMs": 158.59999999403954,
      "totalMs": 11132.699999988079,
      "grade": "good",
      "embeddingsPerSecond": 18.915510719500286
    }
  },
  {
    "precision": "uint8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7959000266393531,
      "negative": 0.42816350597093267,
      "margin": 0.36773652066842044,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 8721.300000011921,
      "inferenceMs": 78.7999999821186,
      "totalMs": 10110.5,
      "grade": "excellent",
      "embeddingsPerSecond": 38.07106599848685
    }
  },
  {
    "precision": "uint8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7951920390125116,
      "negative": 0.42894769632935664,
      "margin": 0.3662443426831549,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 8747,
      "inferenceMs": 157,
      "totalMs": 10133.299999982119,
      "grade": "good",
      "embeddingsPerSecond": 19.10828025477707
    }
  },
  {
    "precision": "q8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.789386296060994,
      "negative": 0.44365669471073066,
      "margin": 0.3457296013502633,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9662.40000000596,
      "inferenceMs": 79,
      "totalMs": 11021.700000017881,
      "grade": "excellent",
      "embeddingsPerSecond": 37.974683544303794
    }
  },
  {
    "precision": "q8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7941647763078531,
      "negative": 0.44616234955631195,
      "margin": 0.34800242675154114,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9407.899999976158,
      "inferenceMs": 150.30000001192093,
      "totalMs": 10891.200000017881,
      "grade": "good",
      "embeddingsPerSecond": 19.96007983873624
    }
  },
  {
    "precision": "q4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.8025921206884732,
      "negative": 0.4634487855493754,
      "margin": 0.3391433351390978,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 13145.59999999404,
      "inferenceMs": 98.40000000596046,
      "totalMs": 14592.199999988079,
      "grade": "excellent",
      "embeddingsPerSecond": 30.487804876202016
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.802592032564747,
      "negative": 0.46344865109644134,
      "margin": 0.33914338146830564,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 12616.600000023842,
      "inferenceMs": 66.69999998807907,
      "totalMs": 14001.09999999404,
      "grade": "excellent",
      "embeddingsPerSecond": 44.9775112524164
    }
  },
  {
    "precision": "q4f16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 11522.399999976158,
      "grade": "failed"
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/graph/graph_utils.cc:30 int onnxruntime::graph_utils::GetIndexFromName(const Node &, const std::string &, bool) itr != node_args.end() was false. Attempting to get index by a name which does not exist:InsertedPrecisionFreeCast_/embeddings/LayerNorm/Constant_output_0for node: /embeddings/LayerNorm/Mul/SimplifiedLayerNormFusion/\n"
  },
  {
    "precision": "q4f16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6466405016170046,
      "negative": 0.477640892871366,
      "margin": 0.16899960874563857,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9663.59999999404,
      "inferenceMs": 60.900000005960464,
      "totalMs": 10985.699999988079,
      "grade": "excellent",
      "embeddingsPerSecond": 49.26108373902104
    }
  },
  {
    "precision": "bnb4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.8010671572276266,
      "negative": 0.4567326450450716,
      "margin": 0.34433451218255495,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 12497.700000017881,
      "inferenceMs": 86.7999999821186,
      "totalMs": 13940.90000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 34.56221198868687
    }
  },
  {
    "precision": "bnb4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.8010672071515009,
      "negative": 0.4567325524965178,
      "margin": 0.34433465465498314,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 12225.59999999404,
      "inferenceMs": 157.7999999821186,
      "totalMs": 13678.40000000596,
      "grade": "good",
      "embeddingsPerSecond": 19.011406846260776
    }
  }
]

BGE Small Chinese v1.5

16 configurations tested
Passed14
Failed2
TestedSep 15, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

Finite normalized distinct 512d vectors; relevant passage cosine exceeds unrelated passage by >=0.1.

PrecisionDeviceResult
FP32GPU · WebGPUAcceleratedPassedVery fast
Q4F16GPU · WebGPUAcceleratedPassedVery fast
FP16GPU · WebGPUAcceleratedPassedVery fast
Q4GPU · WebGPUAcceleratedPassedVery fast
Q8GPU · WebGPUAcceleratedPassedFast
INT8GPU · WebGPUAcceleratedPassedFast
UINT8GPU · WebGPUAcceleratedPassedFast
BNB4GPU · WebGPUAcceleratedPassedFast
FP32CPU · WASMPassedVery fast
Q8CPU · WASMPassedVery fast
UINT8CPU · WASMPassedVery fast
INT8CPU · WASMPassedVery fast
BNB4CPU · WASMPassedVery fast
Q4CPU · WASMPassedVery fast
FP16CPU · WASMFailed—
Q4F16CPU · WASMFailed—
Test setup & sample details
Quality gate
Finite normalized distinct 512d vectors; relevant passage cosine exceeds unrelated passage by >=0.1.
Tested
2026-09-15T20:18:52.108Z
Environment
Isolated headless Chrome 151 on macOS; Transformers.js 4.2.0
Runtime
Transformers.js 4.2.0
Artifact revision
75c43b069aac4d136ba6bc1122f995fedcfd2781
Fixture
{
  "file": "bge-small-zh-v1.5.json",
  "url": "https://f005.backblazeb2.com/file/hubters-webai-public/assets/benchmarks/bge-small-zh-v1.5.json",
  "size": 596,
  "sha1": "2365b808a87dc8e3e77199ae98192506c7826988",
  "source": "WebAI-authored English/Chinese retrieval examples with upstream BGE query instruction",
  "license": "MIT",
  "preprocessing": "Caller supplies query instruction; tokenizer truncation at 512 tokens; normalized CLS pooling."
}
Rerun when
artifact revision change; runtime version change; fixture or quality-gate change; browser/backend behavior change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.664734940397654,
      "negative": 0.23418967951380423,
      "margin": 0.4305452608838498,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 16565,
      "inferenceMs": 73.2999999821186,
      "totalMs": 17929.40000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 40.92769441653266
    }
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6647348151720883,
      "negative": 0.2341897429328582,
      "margin": 0.4305450722392301,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 15941.799999982119,
      "inferenceMs": 58.30000001192093,
      "totalMs": 17265,
      "grade": "excellent",
      "embeddingsPerSecond": 51.45797597575597
    }
  },
  {
    "precision": "fp16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 11480,
      "grade": "failed"
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/graph/graph_utils.cc:30 int onnxruntime::graph_utils::GetIndexFromName(const Node &, const std::string &, bool) itr != node_args.end() was false. Attempting to get index by a name which does not exist:InsertedPrecisionFreeCast_/encoder/layer.2/output/LayerNorm/Constant_output_0for node: /embeddings/LayerNorm/Mul/SimplifiedLayerNormFusion/\n"
  },
  {
    "precision": "fp16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6636284739127313,
      "negative": 0.23314508040540427,
      "margin": 0.430483393507327,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 10776.600000023842,
      "inferenceMs": 59.69999998807907,
      "totalMs": 12161.40000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 50.25125629144123
    }
  },
  {
    "precision": "int8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6735499959857177,
      "negative": 0.2849708610424384,
      "margin": 0.38857913494327934,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 8020.5,
      "inferenceMs": 77.59999999403954,
      "totalMs": 9412.90000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 38.65979381740245
    }
  },
  {
    "precision": "int8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.674504970025274,
      "negative": 0.27837653263899,
      "margin": 0.396128437386284,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 7441.9000000059605,
      "inferenceMs": 121.30000001192093,
      "totalMs": 8773.59999999404,
      "grade": "excellent",
      "embeddingsPerSecond": 24.73206924736332
    }
  },
  {
    "precision": "uint8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6611414559283509,
      "negative": 0.24733703521637118,
      "margin": 0.41380442071197976,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 8135.699999988079,
      "inferenceMs": 77,
      "totalMs": 9459.40000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 38.96103896103896
    }
  },
  {
    "precision": "uint8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6593801901959336,
      "negative": 0.2426119071677546,
      "margin": 0.41676828302817903,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 7852.899999976158,
      "inferenceMs": 125,
      "totalMs": 9312.200000017881,
      "grade": "excellent",
      "embeddingsPerSecond": 24
    }
  },
  {
    "precision": "q8",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.680512248820225,
      "negative": 0.26290890512177784,
      "margin": 0.41760334369844715,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 7102.200000017881,
      "inferenceMs": 76.69999998807907,
      "totalMs": 8379.100000023842,
      "grade": "excellent",
      "embeddingsPerSecond": 39.11342895001654
    }
  },
  {
    "precision": "q8",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6805123651591642,
      "negative": 0.2629088280720719,
      "margin": 0.41760353708709236,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 7877.600000023842,
      "inferenceMs": 120.39999997615814,
      "totalMs": 9266.399999976158,
      "grade": "excellent",
      "embeddingsPerSecond": 24.916943526528787
    }
  },
  {
    "precision": "q4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7217179978917979,
      "negative": 0.33331707157291407,
      "margin": 0.38840092631888384,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 10307.700000017881,
      "inferenceMs": 87.5,
      "totalMs": 11748.300000011921,
      "grade": "excellent",
      "embeddingsPerSecond": 34.285714285714285
    }
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7217179981063002,
      "negative": 0.3333168879104793,
      "margin": 0.3884011101958209,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 10079.90000000596,
      "inferenceMs": 61.099999994039536,
      "totalMs": 11655.799999982119,
      "grade": "excellent",
      "embeddingsPerSecond": 49.09983633866871
    }
  },
  {
    "precision": "q4f16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 9718.5,
      "grade": "failed"
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/graph/graph_utils.cc:30 int onnxruntime::graph_utils::GetIndexFromName(const Node &, const std::string &, bool) itr != node_args.end() was false. Attempting to get index by a name which does not exist:InsertedPrecisionFreeCast_/embeddings/LayerNorm/Constant_output_0for node: /embeddings/LayerNorm/Mul/SimplifiedLayerNormFusion/\n"
  },
  {
    "precision": "q4f16",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.7219709927122462,
      "negative": 0.33403981471764194,
      "margin": 0.38793117799460425,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 8376.299999982119,
      "inferenceMs": 59.20000001788139,
      "totalMs": 9788.90000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 50.67567566036906
    }
  },
  {
    "precision": "bnb4",
    "device": "wasm",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6733756183338517,
      "negative": 0.2764531744589722,
      "margin": 0.3969224438748795,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 11136.699999988079,
      "inferenceMs": 81.59999999403954,
      "totalMs": 12468.40000000596,
      "grade": "excellent",
      "embeddingsPerSecond": 36.76470588503842
    }
  },
  {
    "precision": "bnb4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "shape": true,
      "finite": true,
      "normalized": true,
      "distinct": true,
      "positive": 0.6733755009394152,
      "negative": 0.27645322688476165,
      "margin": 0.3969222740546536,
      "minimumMargin": 0.1,
      "grade": "semantic-retrieval-pass"
    },
    "runtime": {
      "loadMs": 9952.200000017881,
      "inferenceMs": 128.59999999403954,
      "totalMs": 11358.09999999404,
      "grade": "excellent",
      "embeddingsPerSecond": 23.328149301236753
    }
  }
]

Something not quite right?

Tell us what happened with BGE 1.5. Loading, results, or the page itself—we’re here for all of it.

Attachments Optional

Up to 3 files · 50 MB each

Browser, OS, CPU/GPU, memory, screen, storage, and network details are included automatically. Model inputs, results, and browsing history are not collected.