Spaces:
Runtime error
Error converting mistralai/Mistral-7B-Instruct-v0.1
Conversion Settings:
Model: mistralai/Mistral-7B-Instruct-v0.1
Task: None
Compute Units: None
Precision: None
Tolerance: None
Push to: None
Error: "mistral is not supported yet. Only ['bart', 'beit', 'bert', 'big_bird', 'bigbird_pegasus', 'blenderbot', 'blenderbot_small', 'bloom', 'convnext', 'ctrl', 'cvt', 'data2vec', 'distilbert', 'ernie', 'gpt2', 'gpt_bigcode', 'gptj', 'gpt_neo', 'gpt_neox', 'levit', 'llama', 'm2m_100', 'marian', 'mobilebert', 'mobilevit', 'mvp', 'pegasus', 'plbart', 'roberta', 'roformer', 'segformer', 'splinter', 'squeezebert', 't5', 'vit', 'yolos'] are supported. If you want to support mistral please propose a PR or open up an issue."
Hi
@eastriver
! Mistral support is being added to exporters here: https://github.com/huggingface/exporters/pull/62
I'll let you know when the converter supports it :)
I just updated this conversion space to support Mistral. Please @eastriver let us know if it works for you :)
“mistral is not supported yet. Only [‘bart’, ‘beit’, ‘bert’, ‘big_bird’, ‘bigbird_pegasus’, ‘blenderbot’, ‘blenderbot_small’, ‘bloom’, ‘convnext’, ‘ctrl’, ‘cvt’, ‘data2vec’, ‘distilbert’, ‘ernie’, ‘gpt2’, ‘gpt_bigcode’, ‘gptj’, ‘gpt_neo’, ‘gpt_neox’, ‘levit’, ‘llama’, ‘m2m_100’, ‘marian’, ‘mobilebert’, ‘mobilevit’, ‘mvp’, ‘pegasus’, ‘plbart’, ‘roberta’, ‘roformer’, ‘segformer’, ‘splinter’, ‘squeezebert’, ‘t5’, ‘vit’, ‘yolos’] are supported. If you want to support mistral please propose a PR or open up an issue.”
Hi! Thank you for your work,
@pcuenq
.
However, unfortunately, at least the conversion through https://huggingface.co/spaces/coreml-projects/transformers-to-coreml still doesn't work.
I checked:
mistralai/Mistral-7B-v0.1
mistralai/Mistral-7B-Instruct-v0.1
HuggingFaceH4/zephyr-7b-alpha
HuggingFaceH4/zephyr-7b-beta
Hello
@eastriver
! My bad, I did not update the code properly, sorry about that. It's fixed now, so Mistral and Zephyr models are recognized. However, I did a test conversion of Zephyr and got an error at the end of the process ("Error in declaring network."), which is not happening when converting locally in my system. I'm investigating what could be the reason.