cjpais/Handy · error · anyhow::Error

Failed to load gigaam model {}: {}

Error message

Failed to load gigaam model {}: {}

What it means

Thrown when GigaAMModel::load fails to instantiate the Int8 GigaAM ONNX engine from the model files at model_path. The underlying transcribe-rs/ort error indicates the artifacts could not be turned into a session — missing or truncated files, incompatible artifact version, or an execution-provider/memory failure. Handy attaches the model_id, emits the loading_failed model-state-changed event, and leaves no engine loaded.

Source

Thrown at src-tauri/src/managers/transcription.rs:662

                        anyhow::anyhow!(error_msg)
                    })?;
                LoadedEngine::MoonshineStreaming(engine)
            }
            EngineType::SenseVoice => {
                let engine =
                    SenseVoiceModel::load(&model_path, &Quantization::Int8).map_err(|e| {
                        let error_msg =
                            format!("Failed to load SenseVoice model {}: {}", model_id, e);
                        emit_loading_failed(&error_msg);
                        anyhow::anyhow!(error_msg)
                    })?;
                LoadedEngine::SenseVoice(engine)
            }
            EngineType::GigaAM => {
                let engine = GigaAMModel::load(&model_path, &Quantization::Int8).map_err(|e| {
                    let error_msg = format!("Failed to load gigaam model {}: {}", model_id, e);
                    emit_loading_failed(&error_msg);
                    anyhow::anyhow!(error_msg)
                })?;
                LoadedEngine::GigaAM(engine)
            }
            EngineType::Canary => {
                let engine = CanaryModel::load(&model_path, &Quantization::Int8).map_err(|e| {
                    let error_msg = format!("Failed to load canary model {}: {}", model_id, e);
                    emit_loading_failed(&error_msg);
                    anyhow::anyhow!(error_msg)
                })?;
                LoadedEngine::Canary(engine)
            }
            EngineType::Cohere => {
                let engine = CohereModel::load(&model_path, &Quantization::Int8).map_err(|e| {
                    let error_msg = format!("Failed to load cohere model {}: {}", model_id, e);
                    emit_loading_failed(&error_msg);
                    anyhow::anyhow!(error_msg)
                })?;
                LoadedEngine::Cohere(engine)

View on GitHub (pinned to 98a4d80cce)

Solutions

1. Remove and re-download the GigaAM model so ModelManager restores complete artifacts
2. Check the model directory contents and file sizes against the model registry
3. Verify disk space and AV quarantine logs, then retry the download
4. Test with another model to confirm the runtime path is otherwise healthy
5. Update Handy so bundled runtime and model artifacts match

Example fix

// before
let engine = GigaAMModel::load(&model_path, &Quantization::Int8);

// after — guard the load with a size check and actionable error
let onnx_total: u64 = std::fs::read_dir(&model_path)?
    .filter_map(|e| e.ok())
    .map(|e| e.metadata().map(|m| m.len()).unwrap_or(0))
    .sum();
anyhow::ensure!(onnx_total > 0, "GigaAM artifacts missing in {} — re-download the model", model_path.display());
let engine = GigaAMModel::load(&model_path, &Quantization::Int8)?;

// Before loading GigaAM, sanity-check the pack size against the registry expectation
let path = model_manager.get_model_path(model_id)?;
let expected = model_manager.get_model_info(model_id).map(|i| i.size_bytes).unwrap_or(0);
let actual: u64 = std::fs::read_dir(&path)?.filter_map(|e| e.ok())
    .map(|e| e.metadata().map(|m| m.len()).unwrap_or(0)).sum();
anyhow::ensure!(expected == 0 || actual >= expected, "GigaAM incomplete: {} of {} bytes", actual, expected);

match tm.load_model(&model_id) {
    Err(e) if e.to_string().contains("Failed to load gigaam model") => {
        model_manager.remove_local_model(&model_id)?;
        tm.load_model(&model_id)?
    }
    other => other?,
}

When it happens

Trigger: load_model() hits EngineType::GigaAM and GigaAMModel::load(&model_path, &Quantization::Int8) errors: model dir missing the GigaAM .onnx files, partial download, ort provider/dependency missing, or not enough memory to hold the session.

Common situations: Download interrupted by network drop or app exit; disk nearly full; security software removing .onnx files; Handy update bundling a transcribe-rs version whose artifact format changed; small-RAM machines or many GPU apps competing for memory.

Related errors

• Failed to load moonshine model {}: {}
• Failed to load moonshine streaming model {}: {}
• Failed to load SenseVoice model {}: {}
• Failed to load canary model {}: {}
• Failed to load cohere model {}: {}

AI-assisted analysis of cjpais/Handy@98a4d80cce (2026-08-16). Data as JSON: /api/errors/256e002bb5044ffb. Report an issue: GitHub.