transformers
🚨 MXFP4 on the fine-grained machinery, and DeepSeek-V4's FP8 + MXFP4 plan
#49778
Open

🚨 MXFP4 on the fine-grained machinery, and DeepSeek-V4's FP8 + MXFP4 plan #49778

IlyasMoutawwakil wants to merge 15 commits into fg-split/5-fp8 from fg-split/6-mxfp4
IlyasMoutawwakil
IlyasMoutawwakil IlyasMoutawwakil force pushed from 0e0531aa to ae62b387 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 15bcb879 to 551185c0 28 days ago
HuggingFaceDocBuilderDev
IlyasMoutawwakil IlyasMoutawwakil force pushed from ae62b387 to 5b3c9106 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 551185c0 to 4a0d3717 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 5b3c9106 to f41b6f51 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 4a0d3717 to 1f9a2f66 28 days ago
IlyasMoutawwakil IlyasMoutawwakil changed the title MXFP4 on the fine-grained machinery, with DeepSeek-V4's FP8 + MXFP4 plan MXFP4 on the fine-grained machinery, and DeepSeek-V4's FP8 + MXFP4 plan 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from f41b6f51 to d0129d79 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 1f9a2f66 to 909f2cec 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from d0129d79 to b1979e6c 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 909f2cec to 50cb07cc 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from b1979e6c to 4991a3bb 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 4991a3bb to b6b30744 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 8d50a391 to ef5e58c2 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from b6b30744 to d1069758 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from ef5e58c2 to c39d6ae5 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from d1069758 to 3837fcac 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 3837fcac to a1dd17d2 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from c39d6ae5 to cd37228f 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from cd37228f to bc2c9912 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from a1dd17d2 to 7183af2b 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from bc2c9912 to e3a43ca1 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 7183af2b to 75e78c65 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 75e78c65 to c81614cc 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from e3a43ca1 to 32b23c9f 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 32b23c9f to e3a89a2a 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from c81614cc to e4ac490c 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from e4ac490c to 941661e2 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from e3a89a2a to b802f37b 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from b802f37b to 5df081e0 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 941661e2 to b1939494 28 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 5df081e0 to a807b250 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from b1939494 to c3d0d8b2 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from a807b250 to 3acb4399 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from c3d0d8b2 to 15ee3366 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 3acb4399 to 35113a47 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 15ee3366 to 9c003afe 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 35113a47 to f3394183 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 9c003afe to f8f9371e 27 days ago
IlyasMoutawwakil IlyasMoutawwakil changed the title MXFP4 on the fine-grained machinery, and DeepSeek-V4's FP8 + MXFP4 plan 🚨 MXFP4 on the fine-grained machinery, and DeepSeek-V4's FP8 + MXFP4 plan 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from f8f9371e to d9efb5bb 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from f3394183 to 5d804e34 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from d9efb5bb to 0eb9021a 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 0eb9021a to 1aafb732 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from e0fa2b7c to 507c0d36 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 0eb9021a to 1aafb732 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from 1aafb732 to ba15098d 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force-pushed the fg-split/5-fp8 branch from 507c0d36 to 2eb35389 27 days ago
IlyasMoutawwakil IlyasMoutawwakil force pushed from ba15098d to 0b94bcb1 25 days ago
IlyasMoutawwakil Move Mxfp4Config onto the fine-grained machinery, with DeepSeek-V4's …
1c1d6fac
IlyasMoutawwakil Document MXFP4 on the fine-grained machinery and DeepSeek-V4's plan
8c4e1312
IlyasMoutawwakil Test that Mxfp4Config takes main's positional arguments
aa7faff3
IlyasMoutawwakil Require compute capability 8.9 for every MXFP4 activation format
e4e0ee00
IlyasMoutawwakil Dequantize MXFP4 experts for use_kernels on every device
3b1979e5
IlyasMoutawwakil Explain why MXFP4 defaults to DeepSeek's layout and W4A4 experts
4d1cc69d
IlyasMoutawwakil Keep DeepGEMM grouped experts out of DeepSeek-V4's SM100 sweeps
afeff1c8
IlyasMoutawwakil Keep DeepSeek-V4's Triton sweeps on Triton linears through prefer_dee…
acfa6992
IlyasMoutawwakil Read GPT-OSS's experts keys through get_weight_conversions
ea7a1ece
IlyasMoutawwakil Keep main's MXFP4 integration for GPUs below compute capability 8.9
0411198f
IlyasMoutawwakil Read DeepSeek-V4's MXFP4 experts plan in the fp8 reader
6c840c54
IlyasMoutawwakil Leave MXFP4's ModelOpt layout to the producer readers' PR
3e3b270e
IlyasMoutawwakil Give the DeepSeek-V4 load-path fixture its plan explicitly
c3e232f5
IlyasMoutawwakil Restack on the simpler plan resolution and merging
b0750fd3
IlyasMoutawwakil IlyasMoutawwakil force pushed from 0b94bcb1 to b0750fd3 25 days ago
IlyasMoutawwakil Merge remote-tracking branch 'origin/fg-split/5-fp8' into fg-split/6-…
3ec1a724
github-actions

Login to write a write a comment.

Login via GitHub

Reviewers
No reviews
Assignees
No one assigned
Labels
Milestone