diff --git a/docs/opengradient/client/alpha.md b/docs/opengradient/client/alpha.md index d4a8dbd..dd2ae14 100644 --- a/docs/opengradient/client/alpha.md +++ b/docs/opengradient/client/alpha.md @@ -2,7 +2,7 @@ outline: [2,4] --- -[opengradient](../index) / [client](./index) / alpha +B[opengradient](../index) / [client](./index) / alpha # Package opengradient.client.alpha @@ -177,4 +177,76 @@ ModelOutput: The inference result from the contract #### Variables * `inference_abi` : dict -* `precompile_abi` : dict \ No newline at end of file +* `precompile_abi` : dict + + + + + +--- + +## Error Handling + +When calling `infer()`, several errors can occur. Here's how to handle them properly: + +### Basic Error Handling + +```python +from opengradient import Alpha + +alpha = Alpha(private_key='your_private_key') + +try: + result = alpha.infer( + model_cid='QmbUqS93...', + model_input={"data": "input_value"}, + inference_mode=InferenceMode.VANILLA + ) + print(f"Inference succeeded: {result.model_output}") + +except ValueError as e: + # Invalid model CID or model input + print(f"Invalid input: {e}") + +except RuntimeError as e: + # Inference execution failed + print(f"Inference failed: {e}") + +except Exception as e: + # Network error, timeout, insufficient balance, etc. + print(f"Unexpected error: {e}") + + +try: + result = alpha.infer( + model_cid='QmbUqS93...', + model_input={...}, + inference_mode=InferenceMode.VANILLA, + max_retries=3 + ) + +except ValueError as e: + print("Invalid model CID or input format") + # Check that model_cid is a valid IPFS CID + +except RuntimeError as e: + if "Insufficient balance" in str(e): + print("Insufficient wallet balance for inference") + # Fund your wallet with OPG tokens + elif "Timeout" in str(e): + print("Inference timed out - trying again with max_retries") + # Function will auto-retry up to max_retries times + else: + print(f"Inference failed: {e}") + +except Exception as e: + print(f"Unexpected error: {e}") + + +Common Errors and Solutions +Error Cause Solution +ValueError: Invalid model CID Model CID format incorrect Verify CID is valid IPFS format +RuntimeError: Insufficient balance Not enough OPG tokens Fund wallet at opengradient.ai +RuntimeError: Timeout Model taking too long Increase max_retries or check TEE status +RuntimeError: Network error Connection issue Check internet, retry with backoff +RuntimeError: Inference failed Generic inference error Check model_input matches model requirements