Edit model card

bigstral-12b-32k-8xMoE

Made using mergekit MoE branch with the following config:

base_model: abacusai/bigstral-12b-32k
gate_mode: random 
dtype: bfloat16
experts_per_token: 2
experts:
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
  - source_model: abacusai/bigstral-12b-32k
    positive_prompts: []
Downloads last month
1
Safetensors
Model size
81.5B params
Tensor type
BF16
·
Inference API
Input a message to start chatting with bartowski/bigstral-12b-32k-8xMoE.
Model is too large to load in Inference API (serverless). To try the model, launch it on Inference Endpoints (dedicated) instead.

Finetuned from