Training failed. Failed to create trained image after successful training run.

[khenzo]

Hi,
after a pair of successful training (SDXL), all my trainings are ending with this message: "Training failed. Failed to create trained image after successful training run."

I've tried to create a new destination model and create a new API Token to start the creation using the Nodejs Client Library.
I've started a dozen of trainings using the web interface but the response is always the same.

Can someone please have a look into this?

Thank you very much!

[mattt]

Hi @khenzo. Can you please share a link to a failed training? Any code you can share would also be helpful for understanding what the problem might be.

[mattt]

Closing this for now. Happy to reopen if you can share more information to troubleshoot this.

[MaryBIZCO]

HI
I HAVE THE SAME PROBLEM
ANY ADVICE PLEASE

[MaryBIZCO]

Training failed.

Prediction input failed validation: {"detail":[{"loc":["body","input","images"],"msg":"'' is not a valid URL scheme. 'data', 'http', or 'https' is supported.","type":"value_error"}]}

[MaryBIZCO]

{
"completed_at": "2024-10-20T18:23:35.925477Z",
"created_at": "2024-10-20T18:23:35.920000Z",
"data_removed": false,
"error": "Prediction input failed validation: {"detail":[{"loc":["body","input","images"],"msg":"'' is not a valid URL scheme. 'data', 'http', or 'https' is supported.","type":"value_error"}]}",
"id": "9es8e34zy1rj20cjncns14fny4",
"input": {
"steps": 1000,
"images": "marymway",
"repo_id": "marymway/flux.1-dev-lora",
"hf_token": "[REDACTED]",
"batch_size": 1,
"model_name": "marymway/flux.1-dev",
"resolution": "512,768,1024",
"lora_linear": 16,
"learning_rate": 0.0004,
"lora_linear_alpha": 16
},
"logs": null,
"metrics": {
"total_time": 0.005477
},
"output": null,
"started_at": "2024-10-20T18:23:35.925477Z",
"status": "failed",
"urls": {
"get": "https://api.replicate.com/v1/trainings/9es8e34zy1rj20cjncns14fny4",
"cancel": "https://api.replicate.com/v1/trainings/9es8e34zy1rj20cjncns14fny4/cancel"
},
"version": "06e50f60983aa9ad9e4c13ba1d56ee235925ee6bd1c604d94c26a3322aeb8d47"
}