mlc-ai/mlc-llm

Phi-3 mini 4k instruct with MICROSOFT's quantization

Open

#2273 aperta il 4 mag 2024

Vedi su GitHub
 (3 commenti) (0 reazioni) (0 assegnatari)Python (1220 fork)batch import
help wantednew-models

Metriche repository

Star
 (16.227 star)
Metriche merge PR
 (Merge medio 4g) (2 PR mergiate in 30 g)

Descrizione

⚙️ Request New Models

Additional context

I know others have made this request already (https://github.com/mlc-ai/mlc-llm/issues/2246, https://github.com/mlc-ai/mlc-llm/pull/2222, https://github.com/mlc-ai/mlc-llm/issues/2238, https://github.com/mlc-ai/mlc-llm/issues/2205).

But I am requesting something different: I am suggesting that you do not quantize or modify the weights of the model but that you instead use Microsoft's already 4-bit quantized weights.

The reason is that I suspect (although it is not explicit in their repo) they used quantization-aware training to build these GGUF files. I have tested the regular 32-bit model vs the GGUF 4-bit one and the performance is almost equivalent which is not what I've seen so far with MLC's quantized models (they tend to be more inaccurate compared to their 32-bit counterparts).

Is there a way to use Microsoft's own quantized weights?

Thank you! Federico

Guida contributor