ml-commons
ml-commons copied to clipboard
Fine tune error message for model content changed
When user upload model via URL and didn't provide model hash value, it will throw "model content changed" error, which is hard to understand. We should clearly tell user that model content has value not set.
POST /_plugins/_ml/models/_upload
{
"name": "all-MiniLM-L6-v2",
"version": "1.0.0",
"description": "test model",
"model_format": "TORCH_SCRIPT",
"model_config": {
"model_type": "bert",
"embedding_dimension": 384,
"framework_type": "sentence_transformers"
},
"url": "https://github.com/opensearch-project/ml-commons/raw/2.x/ml-algorithms/src/test/resources/org/opensearch/ml/engine/algorithms/text_embedding/all-MiniLM-L6-v2_torchscript_sentence-transformer.zip?raw=true"
}
Task
{
"task_type": "REGISTER_MODEL",
"function_name": "TEXT_EMBEDDING",
"state": "FAILED",
"worker_node": [
"_QJb--HRS2-7lfq5DCWMiQ"
],
"create_time": 1693505198395,
"last_update_time": 1693505199947,
"error": "model content changed",
"is_async": true
}
Need to calcuate sha256 first
shasum -a 256 all-MiniLM-L6-v2_torchscript_sentence-transformer.zip
Then upload
POST /_plugins/_ml/models/_register
{
"name": "all-MiniLM-L6-v2",
"version": "1.0.0",
"description": "test model",
"model_format": "TORCH_SCRIPT",
"model_group_id": "FTNlQ4gBYW0Qyy5ZoxfR",
"model_content_hash_value": "c15f0d2e62d872be5b5bc6c84d2e0f4921541e29fefbef51d59cc10a8ae30e0f",
"model_config": {
"model_type": "bert",
"embedding_dimension": 384,
"framework_type": "sentence_transformers"
},
"url": "https://artifacts.opensearch.org/models/ml-models/huggingface/sentence-transformers/all-MiniLM-L6-v2/1.0.1/torch_script/sentence-transformers_all-MiniLM-L6-v2-1.0.1-torch_script.zip"
}
@ylwu-amzn can you assign me to this issue? Thank you
Sure, thanks @TrungBui59, assigned to you
@TrungBui59 , any update ? We plan to fix this in 2.12, if you don't have enough bandwidth, will ask someone else to help.
@ylwu-amzn sorry for the delayed response, I was having some personal issues, so I wasn't checking frequently, I am still working on it but it seems like others have posted a PR for this issue