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.

Sweep Next-Edit 1.5B — ONNX

Predict an updated code file from repository context, recent changes, and the file's contents. Returns a complete revised file. Q4 WebGPU is verified on the recorded code-edit fixture; review generated code before applying it.

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

Predict an updated code file from repository context, recent changes, and the file's contents. Returns a complete revised file. Q4 WebGPU is verified on the recorded code-edit fixture; review generated code before applying it.

Created by
Sweep AI
License
Apache-2.0
Model released
Unknown
Source repository created
Jan 20, 2026
Model weights · Q4
943.3 MB

Best for

  • Predicting a likely next code edit from a developer's recent changes.
  • Prototyping local code-edit autocomplete in browser developer tools and editors.
  • Benchmarking causal language-model ONNX artifacts with the reusable Transformers.js LLM harness.

Keep in mind

  • This is a specialized next-edit predictor, not a general conversational assistant.
  • Predictions may be incorrect, incomplete, insecure, or destructive and require developer review before application.
  • Quality depends on exact prompt formatting and useful repository and edit context; unrelated or truncated context can substantially reduce accuracy.
  • The model can reproduce sensitive or licensed material supplied in its context; callers are responsible for reviewing inputs and outputs.
  • The verified Q4 download is approximately 946 MB before browser runtime assets and may exceed memory or cache limits on mobile, private browsing, or constrained devices.

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
sweepai/sweep-next-edit-1.5B
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
Jan 22, 2026
Browser model files
Xenova/sweep-next-edit-1.5B
Files updated
Jan 24, 2026
Artifact revision
402c437f67cee450c00d3add669b43ae79e55e50

Input

TEXT

Sample result

Sweep Next-Edit 1.5B — ONNX

Settings

Maximum number of tokens in the predicted updated file. Lower values finish sooner but can cut answers short. Higher values allow longer answers and can use more time and memory; they do not guarantee better answers.
Sampling candidate count. The listed range is a frontend selection policy, not a claimed model or runtime maximum; the adapter may accept values outside it. Lower values show fewer top results. Higher values include weaker matches and more output. Automatic/null returns all results when offered; it does not mean the model is more accurate.
Advanced settings 5
Use sampling instead of deterministic greedy decoding. On allows varied responses and makes sampling controls relevant. Off favors the most likely tokens for more repeatable responses; it does not guarantee correctness.
Sampling temperature; ignored by greedy decoding. Lower values favor likely tokens and more repeatable wording. Higher values increase variation and can increase mistakes. It affects sampling only; some runtimes use zero for greedy decoding.
Nucleus-sampling probability mass. Lower values restrict word choice to more likely tokens. Higher values allow more varied but potentially less reliable wording. It affects sampling only; keep the default unless tuning generation.
Penalty applied to repeated tokens. The listed range is a frontend selection policy, not a claimed model or runtime maximum; the adapter may accept values outside it. 1 leaves scores unchanged. Values above 1 discourage repetition but can distort wording if too high. Values below 1 favor previously used tokens and can increase repetition.
Remove tokenizer control tokens from the returned edit. On removes special tokens for readable text. Off exposes control markers for inspection and can make output unsuitable for direct display.

Output

Your result will appear here

Add input or try the sample, then run the model.

Benchmark evidence

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

Sweep Next-Edit 1.5B — ONNX

10 configurations tested
Passed1
Failed9
TestedNov 4, 2026
RuntimeTransformers.js 4.2.0

Quality criteria

No explicit quality gate recorded. Review per-case evidence; full certification is pending.

PrecisionDeviceResult
Q4GPU · WebGPUAcceleratedPassedModerate
FP32CPU · WASMFailed—
FP32GPU · WebGPUFailed—
FP16CPU · WASMFailed—
FP16GPU · WebGPUFailed—
Q8CPU · WASMFailed—
Q8GPU · WebGPUFailed—
Q4F16CPU · WASMFailed—
Q4F16GPU · WebGPUFailed—
Q4CPU · WASMFailed—
Test setup & sample details
Quality gate
No explicit quality gate recorded. Review the per-case quality evidence; full certification is pending.
Tested
2026-11-04T19:18:16.527Z
Environment
Chrome 151.0.7922.176 on macOS arm64; Transformers.js 4.2.0
Runtime
Transformers.js 4.2.0
Artifact revision
402c437f67cee450c00d3add669b43ae79e55e50
Fixture
{
  "file": "sweep-next-edit-1.5b.json",
  "url": "https://f005.backblazeb2.com/file/hubters-webai-public/assets/benchmarks/sweep-next-edit-1.5b.json",
  "size": 1291,
  "sha1": "98a29ff43b6bdcd2faba07f9ba47bb9d9ed36192"
}
Rerun when
model artifact revision change; Transformers.js or ONNX Runtime version change; fixture or quality-gate change; browser/backend behavior change
Per-case quality & failure details
[
  {
    "precision": "fp32",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7676,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: Deserialize tensor model.layers.7.input_layernorm.weight failed.Failed to load external data file \"\"model.onnx_data\"\", error: Module.MountedFiles is not available."
  },
  {
    "precision": "fp32",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7366.100000023842,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: Deserialize tensor model.layers.11.mlp.up_proj.MatMul.weight failed.Failed to load external data file \"\"model.onnx_data_1\"\", error: Module.MountedFiles is not available."
  },
  {
    "precision": "fp16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7499.699999928474,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/optimizer/initializer.cc:51 onnxruntime::Initializer::Initializer(const onnx::TensorProto &, const std::filesystem::path &) [ONNXRuntimeError] : 1 : FAIL : Failed to load external data file \"\"model_fp16.onnx_data\"\", error: Module.MountedFiles is not available.\n"
  },
  {
    "precision": "fp16",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7566.5,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: Deserialize tensor model.layers.14.mlp.up_proj.MatMul.weight failed.Failed to load external data file \"\"model_fp16.onnx_data\"\", error: Module.MountedFiles is not available."
  },
  {
    "precision": "q8",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7987.799999952316,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 9, ERROR_MESSAGE: Could not find an implementation for GatherBlockQuantized(1) node with name '/model/embed_tokens/Gather_Quant'"
  },
  {
    "precision": "q8",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7678.300000071526,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: Deserialize tensor model_layers_14_attn_q_proj_MatMul_weight_zp failed.Failed to load external data file \"\"model_quantized.onnx_data\"\", error: Module.MountedFiles is not available."
  },
  {
    "precision": "q4f16",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7555.699999928474,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: /mnt/vss/_work/1/s/onnxruntime/core/optimizer/initializer.cc:51 onnxruntime::Initializer::Initializer(const onnx::TensorProto &, const std::filesystem::path &) [ONNXRuntimeError] : 1 : FAIL : Failed to load external data file \"\"model_q4f16.onnx_data\"\", error: Module.MountedFiles is not available.\n"
  },
  {
    "precision": "q4f16",
    "device": "webgpu",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 7826.200000047684,
      "grade": "failed",
      "firstChunkMs": null,
      "outputTokens": null,
      "tokensPerSecond": null
    },
    "failure": "Error: Can't create a session. ERROR_CODE: 1, ERROR_MESSAGE: Deserialize tensor model_layers_27_mlp_down_proj_MatMul_weight_zp failed.Failed to load external data file \"\"model_q4f16.onnx_data\"\", error: Module.MountedFiles is not available."
  },
  {
    "precision": "q4",
    "device": "wasm",
    "status": "fail",
    "quality": "Not embedded",
    "runtime": {
      "loadMs": null,
      "inferenceMs": null,
      "totalMs": 153972.90000000596,
      "grade": "failed"
    },
    "failure": "Error: [WorkerError:runtime_error] init: Can't create a session. ERROR_CODE: 9, ERROR_MESSAGE: Could not find an implementation for GatherBlockQuantized(1) node with name '/model/embed_tokens/Gather_Quant'"
  },
  {
    "precision": "q4",
    "device": "webgpu",
    "status": "pass",
    "quality": {
      "passed": true,
      "requiredCodeRetained": true,
      "expectedEdit": true,
      "repeatedLineRatio": 0
    },
    "runtime": {
      "loadMs": 96958.09999999404,
      "inferenceMs": 836.5,
      "totalMs": 97794.59999999404,
      "charactersPerSecond": 141.06395696353854,
      "grade": "good"
    }
  }
]
Unavailable checkpoints (1)

These checkpoints have not passed the required browser checks.

0.5B · Unavailable in this browser runtime: original GGUF weights have no published compatible ONNX artifact for the selected Transformers.js path. A validated conversion or GGUF backend and independent browser quality/device checks are required. Native0.5B quality failures are recorded separately.

Something not quite right?

Tell us what happened with Sweep Next Edit. 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.