Unverified50% confidenceFactExact time
Change #19149 adds DFlash2 support to MLX, an expansion of that backend's support matrix.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/27/2026
First Seen
Valid until: 1/25/2027
Sources
Related Entities
Related Claims
UnverifiedOllama v0.40.4 contains three changes directly related to MLX: state compaction after speculative rollback and an MLX / MLX-C version upgrade, both submitted by dhiltgen.63% similarUnverifiedChange #19291 bumps the version numbers of both MLX and MLX-C, officially described as improving prompt processing speed for the gemma, flint and laguna model families.61% similarUnverifiedvLLM v0.28.0rc2 introduces a new speculative decoding feature called DFlash2, which combines two major technical components: local convolution and a candidate selector (PR #52816).61% similarUnverifiedThe MLX backend of Ollama v0.40.10 supports speculative decoding using a DFlash draft model for the Gemma 4 target model.60% similarUnverifiedMJX's Warp backend combines JAX's automatic differentiation with Warp's GPU kernels, further improving simulation speed.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1101756API
curl https://kongchang.com/api/v1/knowledge/claims/1101756MCP
get_claim(id=1101756)