Embedding Error Snowflake Ollama

[Fau57]

After picking the snowflake embedding model in the options and reloading the vault i get this response: Error: Embedding function error: Error: Resource initialization failed: Error: Pipeline initialization failed for repo Error: local_files_only=true or env.allowRemoteModels=false and file was not found locally at "C:\Users\user.ollama\models\manifests\registry.ollama.ai\library/snowflake-arctic-embed/tokenizer.json". Stack Trace: Error: local_files_only=true or env.allowRemoteModels=false and file was not found locally at "C:\Users\user.ollama\models\manifests\registry.ollama.ai\library/snowflake-arctic-embed/tokenizer.json". at getModelFile (file:///C:/Users/user/AppData/Local/Programs/Reor/resources/app.asar/node_modules/@xenova/transformers/src/utils/hub.js:462:27) at async getModelJSON (file:///C:/Users/user/AppData/Local/Programs/Reor/resources/app.asar/node_modules/@xenova/transformers/src/utils/hub.js:575:18) at async Promise.all (index 0) at async loadTokenizer (file:///C:/Users/user/AppData/Local/Programs/Reor/resources/app.asar/node_modules/@xenova/transformers/src/tokenizers.js:61:18) at async AutoTokenizer.from_pretrained (file:///C:/Users/user/AppData/Local/Programs/Reor/resources/app To Reproduce Steps to reproduce the behavior:

1. Changed model in local embedding model to snowflake-arctic-embed i think its the 33 million model,
2. Click on close program to reinitialize the vault
3. Reopen Reor
4. See above for error information

Expected behavior my understanding is it should re do the embedding with the new model or am i wrong?

Desktop (please complete the following information):

• OS: win 11
  • Hardware: Ryzen 5 5560U and 32 GB ddr4 RAM
• Version: i think its the most reccent

[uuur86]

I guess, Reor only supports "*.onnx" type models.

Also, file structure must contain "tokenizer.json, config.json, tokenizer_config.json" files. The missing files can be downloaded from hf ( https://huggingface.co/Snowflake/snowflake-arctic-embed-m/tree/main )

The directory structure should be as follows for 'C:\Users\user.ollama\models\manifests\registry.ollama.ai\library/snowflake-arctic-embed':

• onnx (dir) -- model.onnx -- model_quantized.onnx
• tokenizer.json
• config.json
• tokenizer_config.json

[Fau57]

Oh i must have missed that in my reading sorry my bad

On Thu, Jun 6, 2024, 10:21 a.m. Uğur Biçer @.***> wrote:

I guess, Reor only supports "*.onnx" type models.

— Reply to this email directly, view it on GitHub https://github.com/reorproject/reor/issues/264#issuecomment-2152925035, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGA3UE2DWJMAGPSDZVK5XQDZGCEARAVCNFSM6AAAAABIXKBKDGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNJSHEZDKMBTGU . You are receiving this because you authored the thread.Message ID: @.***>

[uuur86]

No problem! Just wanted to make sure everything is clear. If you need any further details or assistance, feel free to ask :)

[Fau57]

I'm not sure this is your project or you're just a member but I'm wondering I guess I could just find your roadmap but I was wondering what that looks like

On Thu, Jun 6, 2024, 2:18 p.m. Uğur Biçer @.***> wrote:

No problem! Just wanted to make sure everything is clear. If you need any further details or assistance, feel free to ask :)

— Reply to this email directly, view it on GitHub https://github.com/reorproject/reor/issues/264#issuecomment-2153338998, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGA3UE3J4J5Z4TLA7A3L4L3ZGC7ZLAVCNFSM6AAAAABIXKBKDGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNJTGMZTQOJZHA . You are receiving this because you authored the thread.Message ID: @.***>

[uuur86]

Sorry, I didn't understand what you mean. But;

I'm a volunteer contributor to this project. Open source culture emphasizes collective improvement and shared ownership. "open source culture" <-- You can search this on google for detailed information..

[Fau57]

i am an aspiring developer. I think i am at beginner level for python at least, ANYWAY. I also use the crap outta Obsidian. Im just wondering if there are plans from you guys involved to add more features? or if i should fork and just go to town?

[uuur86]

You can follow the Reor project on Discord. There, you can also suggest features and discuss them. https://discord.gg/b7zanGCTUY

[uuur86]

Additionally, I believe everyone would be happy to welcome aspiring developers like you to join us!