CLIWorks/Spider-FLEXITOKENS-FP8
Spider-FLEXITOKENS FP8 Training FP8 training pipeline for Spider-FLEXITOKENS on NVIDIA Blackwell GPUs (sm_120) using torchao Float8Linear and optional TileKernels fused MoE routing. Architecture Spider is a Recurrent-Depth Transformer (RDT) with: 1B parameters (996M), hidden_size=2048 Byte-level vocab: 272 tokens (256 UTF-8 bytes + 16 specials: BOS=257, EOS=258, PAD=256) 6 recurrent layers with MoE (32 experts, top-2 routing) + MLA attention 2 prelude + 2 coda… See the full description on the dataset page: https://huggingface.co/datasets/CLIWorks/Spider-FLEXITOKENS-FP8.
| Author / organization | CLIWorks |
|---|---|
| License (as declared) | apache-2.0 |
| Downloads | 239 (as reported by Hugging Face) |
| Likes | 0 |
| Created | 10 May 2026 |
| Last updated at source | 10 May 2026 |
Popularity, last 3 days
239 now (-15 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 16 Oct 2026 Description: Spider-FLEXITOKENS FP8 Training FP8 training pipeline for Spider-FLEXITOKENS on NVIDIA Blackwell GPUs (sm_120) using torchao Float8Linear and optional TileKernels fused MoE routing. Architecture Spider is a Recurrent-Depth Transformer (RDT) → Spider-FLEXITOKENS FP8 Training FP8 training pipeline for Spider-FLEXITOKENS on NVIDIA Blackwell GPUs (sm_120) using torchao Float8Linear and optional TileKernels fused MoE routing. Architecture Spider is a Recurrent-Depth Transformer (RDT)
Changes detected by Silytics between refreshes of the source.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.