Downloads · 30 days
236
1% of all-time downloads
saarvajanik/facebook-opt-6.7b-gqa-ub-16-best-for-KV-cache
facebook-opt-6.7b-gqa-ub-16-best-for-KV-cache is a text generation model from saarvajanik. Use it when you need the model to write or continue text. It is set up for transformers. The card lists the license as mit.
This is a QCQA version of the original model facebook/opt-125m. In this version, the original MHA architecture is preserved but instead of having a single K/V head, different K/V heads corresponding to the same group…
Downloads · 30 days
236
1% of all-time downloads
All-time downloads
15.9K
Public
Repo size
82.4 GB
Likes
15
Public
Click a slice to open those files.
.bin27.5 GB · 100%
From the Hugging Face model README
This is a QCQA version of the original model facebook/opt-125m. In this version, the original MHA architecture is preserved but instead of having a single K/V head, different K/V heads corresponding to the same group have the same mean-pooled K or V values. It has 16 groups of KV heads per layer instead of original 32 KV heads in the MHA implementation.