poolside
/

Laguna-M.1-NVFP4

Text Generation

Mixture of Experts

8-bit precision

compressed-tensors

Model card Files Files and versions

Rewrite model card to match Laguna-XS.2-NVFP4 template

#1

by joerowell - opened 22 days ago

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

initial commita1c8bc46

Initial NVFP4 conversion of Laguna-M.1 (private)249d2ea6

Copy model card from Laguna-M.13c89a248

Adapt model card for NVFP4 variant3a81530d

Bump max_position_embeddings to 262144 and YaRN factor to 64 (matches Laguna-M.1)e535a4b0

Drop redundant --dtype bfloat16 from quickstart3ade6185

update sampling parameters to match evalse1329786

Rewrite model card to match Laguna-XS.2-NVFP4 templatecc0b8c58

Poolside org 22 days ago

No description provided.

Add LICENSE.mdb1b94e18

Fix vLLM (#41129) and TRT-LLM (#13559, #15110) references7260b560

Add FP8-KV-cache vLLM >= 0.22.0 caveat (vllm#42650)5ad6a14e

Remove FP8-KV-cache caveat (not applicable to Laguna M.1)f89325b1

joerowell changed pull request status to merged 22 days ago

joerowell deleted the refs/pr/1 ref 22 days ago

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment