Zackriya-Solutions/meetily · error

{} has {} bytes, expected exactly {} bytes

Error message

{} has {} bytes, expected exactly {} bytes

What it means

Thrown by validate_model_directory when an artifact exists but its size on disk does not exactly match the expected exact_bytes from the ArtifactSpec. The Parakeet engine requires byte-exact artifacts, so any truncated, padded, or tampered file is rejected before the model is loaded.

Source

Thrown at frontend/src-tauri/src/parakeet_engine/parakeet_engine.rs:419

                available_models.insert(model.name.clone(), model.clone());
            }
            drop(available_models);
            drop(active_downloads);

            for (model_name, error) in validation_errors {
                log::warn!("Model directory {} appears corrupted: {}", model_name, error);
            }
            return Ok(models);
        }
    }

    fn validate_model_directory(model_dir: &Path, artifacts: &[ArtifactSpec]) -> Result<()> {
        for artifact in artifacts {
            let path = model_dir.join(artifact.filename);
            let metadata = std::fs::metadata(&path)
                .map_err(|error| anyhow!("Failed to read {} metadata: {}", artifact.filename, error))?;
            if metadata.len() != artifact.exact_bytes {
                return Err(anyhow!(
                    "{} has {} bytes, expected exactly {} bytes",
                    artifact.filename,
                    metadata.len(),
                    artifact.exact_bytes
                ));
            }
        }

        Ok(())
    }

    /// Load a Parakeet model
    pub async fn load_model(&self, model_name: &str) -> Result<()> {
        let model_info = {
            let models = self.available_models.read().await;
            models
                .get(model_name)
                .cloned()

View on GitHub (pinned to a2cb62e827)

Solutions

1. Delete the offending artifact (or the whole model directory) and re-download the model through the app so sizes match the spec
2. Ensure downloads are resumed using Content-Range validation so partial files are not accepted
3. Check you downloaded the exact model variant/quantization matching the app's ArtifactSpec
4. Verify available disk space and that no antivirus truncated the file
5. Update the app if the upstream model release changed file sizes (spec must be updated too)

Example fix

// before: accepting whatever file is present
let f = model_dir.join("encoder.onnx"); // 4,998,123 bytes, expected 5,000,000
// after: re-download with size verification
if std::fs::metadata(&f)?.len() != spec.exact_bytes { std::fs::remove_file(&f)?; download_artifact(spec)?; }

fn artifact_size_matches(model_dir: &Path, filename: &str, expected: u64) -> bool { std::fs::metadata(model_dir.join(filename)).map(|m| m.len() == expected).unwrap_or(false) }

match load_parakeet_model(name) { Err(e) if e.to_string().contains("expected exactly") => { delete_model_dir(model_dir)?; download_model(name).await?; load_parakeet_model(name).await }, other => other }

When it happens

Trigger: Post-download validation or model-load check finds e.g. an encoder file of 4,998,123 bytes where the spec requires 5,000,000 — usually from an interrupted/partial download, a truncated transfer, or a version mismatch between the downloaded model release and the artifact spec baked into the app.

Common situations: Network drop during model download left a partial file that later passes a resume check incorrectly; user replaced a model file with a different quantization/variant; model host updated the artifact bytes while the app still expects old sizes; disk-full during write.

Understand the failure class

Background: Checksum mismatch errors: "checksum verification failed", "digest mismatch", "expected vs actual checksum" — what they mean and how to fix them — this error's family across 41 libraries.

Related errors

• Failed to delete incomplete file {}: {}
• Failed to open file for resume {}: {}
• Failed to create file {}: {}
• Failed to write chunk to file: {}
• Failed to open file for append: {}

AI-assisted analysis of Zackriya-Solutions/meetily@a2cb62e827 (2026-09-12). Data as JSON: /api/errors/d95e54fb17dcae22. Report an issue: GitHub.