Support for DeepseekV32ForCausalLM with DeepSeek Sparse Attention (DSA)#21149

Draft

fairydreaming wants to merge 52 commits intoggml-org:masterfrom

fairydreaming:deepseek-dsa

Commits on Mar 12, 2026

model : Initial support for DeepseekV32ForCausalLM (for now with dense attention). Needs manual change of add_bos_token to true in tokenizer_config.json before conversion.
sszymczy
committed
model : added indexer q and k calculation in DeepseekV32ForCausalLM.
sszymczy
committed
ggml : add Hadamard transform GGML OP and implementation
sszymczy
committed

Commits on Mar 13, 2026

kv-cache : add cache for indexer keys (temporary solution)
sszymczy
committed

Commits on Mar 14, 2026

Commits on Mar 15, 2026

Commits on Mar 16, 2026

Commits on Mar 22, 2026

model : replaced ggml_argsort_top_k with ggml_top_k in DeepSeek V3.2 indexer implementation since the former fails for large tensors even when using CCCL.
sszymczy
committed

Commits on Mar 23, 2026

model : added comments in DeepSeek V3.2 lightning indexer implementation.
sszymczy
committed

Commits on Mar 25, 2026

Commits on Mar 31, 2026

Merge remote-tracking branch 'upstream/master' into deepseek-dsa
sszymczy
committed

Commits on Apr 1, 2026

Commits on Apr 8, 2026

ggml : added new GGML_OP_LIGHTNING_INDEXER that merges ggml_mul_mat(), ggml_relu() and ggml_sum_rows() to save compute buffer memory.
sszymczy
committed

Commits on Apr 12, 2026

ggml : support lightning indexer key quantization
sszymczy
committed

Commits on Apr 14, 2026

Commits on Apr 20, 2026