Downloads · 30 days
34
20% of all-time downloads
whale-v3-base-merged
theblackcat102/whale-v3-base-merged
whale-v3-base-merged is a text generation model from theblackcat102. Use it when you need the model to write or continue text. It is set up for transformers.
Since Deepseek v3 dense layers ( first 3 ) happens to be 18432 which equals to 9 x 2048. Since there's 256 experts with 2048 dimensions as the intermediate dimension, we can first average all 32 experts as 1 expert an…
All-time downloads
167
Public
Parameters
37.4B
100 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors100 GB · 100%
Parameter types
How the weights are stored.
BF1624.8B · 66%
Try a prompt
Model card
From the Hugging Face model README
Deepseek-v3-Base Group 8 Average Weights
Since Deepseek v3 dense layers ( first 3 ) happens to be 18432 which equals to 9 x 2048. Since there's 256 experts with 2048 dimensions as the intermediate dimension, we can first average all 32 experts as 1 expert and concatenate all 8 groups into a 16384 layers. Adding the share_experts in to the new MLP layers and we can get 18432 MLP layers.
Model Details
Model Description
Unfortunately, this model doesn't work out of the box ( after dequantize and merging ) all it generates is giberish tokens. So either my code sucks or merging all that experts down breaks the model too much that every brokes.
I'm trying to recover the MLP layers by pretraining, but I'm bit low on compute and doesn't have much to spare. Also if you have small corpus which I can use feel free to comments and suggest what I should do next
QLoRA pretrained test run can be found here : theblackcat102/whale-v3-base-concept-test-lora-380
- Task
- Text Generation
- Library
- transformers
- Type
- deepseek_v3
- Created
- Jan 10, 2025
- Updated
- Jan 12, 2025