> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt.
> For full documentation content in one file, see https://docs.sarvam.ai/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

# Troubleshooting

> Fix common issues with Sarvam SageMaker deployments: failed model creation, endpoints stuck creating, invoke errors, latency and throttling. Diagnose with CloudWatch.

Start with **CloudWatch Logs** for the endpoint — the model container streams its logs there, and most invoke failures explain themselves in the log line for that request.

## Deployment issues

#### create\_model fails immediately

Usually a region mismatch or a permissions gap. Confirm the **model package ARN is for the same region** as your client, that you're **subscribed** to the listing, and that your role has `AmazonSageMakerFullAccess` and `AWSMarketplaceManageSubscriptions`.

#### Endpoint stuck in Creating, then Failed

Almost always a **service quota** for the GPU instance type. Check the failure reason on the endpoint, then request the instance quota in the Service Quotas console and retry.

## Invoke errors

#### 4xx / ModelError on a request

Every non-2xx arrives as a boto3 **`ModelError` (HTTP 424)** — read the real status in `OriginalStatusCode` and the body in `OriginalMessage` (see [Error handling on SageMaker](/api/self-hosted/sagemaker/errors)). Check the `Content-Type`: real-time Saaras v3 needs `multipart/form-data; boundary=…`, Bulbul v3 needs `application/json`, and Vision needs the document's MIME type. Confirm `model` is **`saaras:v3`** or **`bulbul:v3`**, not a package version like `saaras:v3.1`.

#### Request rejected for length or size

Audio over **30 s** on Saaras v3 REST, Bulbul v3 text over **2500 chars** (real-time), a Vision sync document over **\~5 pages** (or over the **500-page** hard cap) or one that can't finish inside the **60 s** `InvokeEndpoint` window, or a payload over **\~6 MB** — move to streaming (Saaras v3 / Bulbul v3) or async / batch (Vision). See [Limitations](/api/self-hosted/sagemaker/limitations).

#### Streaming connection refused

Streaming accepts only **8 kHz or 16 kHz, mono, 16-bit PCM**. Resample the audio and confirm the `sample_rate` query parameter matches the stream.

## Latency & throttling

#### 429 / 503 service\_overloaded from the container

At capacity the container **sheds fast** with **`429`** (Saaras v3) or **`503`** (Bulbul v3, Sarvam Vision) carrying `error.code: "service_overloaded"` and a body `retry_after` — this is expected backpressure and is retryable. Back off for `retry_after` seconds with jitter. It is distinct from SageMaker's own `ThrottlingException` (an invocation-rate limit on the AWS API, not the model). See [Error handling on SageMaker](/api/self-hosted/sagemaker/errors).

#### ThrottlingException or rising latency under load

The endpoint is saturated. Add instances or enable [autoscaling](/api/self-hosted/sagemaker/operations#autoscaling), move to a larger instance, or raise per-instance concurrency (see [Configure & tune](/api/self-hosted/sagemaker/configure)). Watch `ModelLatency` and `InvocationsPerInstance` in CloudWatch.

#### High cost between bursts

Switch spiky workloads to an **async** endpoint with **scale-to-zero**, so you don't pay for idle instances.

Still stuck? Contact [developer@sarvam.ai](mailto:developer@sarvam.ai) with your endpoint name, region, and the relevant CloudWatch log lines.