Carry ggml-org#24423 (DiffusionGemma) merged onto b10630 by danielhanchen · Pull Request #107 · unslothai/llama.cpp

Commits on Jun 10, 2026

diffusion-visual updates
danielhanchen
committed
c5fe75b
diffusion: fix Windows build, skip diffusion-gemma in test-llama-archs, drop debug hooks
danielhanchen
committed
c84e85a

Commits on Jun 11, 2026

diffusion-cli: note that --fit is not applied by the diffusion runner
danielhanchen
committed
7c200dd
diffusion-cli: honor -ot / --n-cpu-moe (copy tensor_buft_overrides into model params)
danielhanchen
committed
9b4beb7
diffusion: device-resident self-conditioning + cli throughput, /help, /clear
danielhanchen
committed
15ad8f4
diffusion: device-side sampling reductions (default on, ~1.4x/step)
danielhanchen
committed
d6cf0b2
ggml-cuda: fix diffusion sampler build on HIP/MUSA
danielhanchen
committed
53752ad
diffusion: add gemma visual server for streaming canvas frames
danielhanchen
committed
f53de16
diffusion: self-tokenize in gemma visual server
danielhanchen
committed
7ea238c
diffusion: stop per-step device-sample fallback spam
danielhanchen
committed
10a2613

Commits on Jun 12, 2026

Commits on Jun 23, 2026

Commits on Jul 3, 2026

Fix merge conflicts
danielhanchen
committed
1d2faac

Commits on Jul 4, 2026

Add tool calling
danielhanchen
committed
c3fb972

Commits on Aug 10, 2026

Fix merge conflicts
danielhanchen
committed
1d87a1b
Fix merge conflicts
danielhanchen
committed
daca807

Commits on Aug 26, 2026

Merge b10630 into the DiffusionGemma pin
danielhanchen
committed
a15ca5b
diffusion-gemma : port the visual server to common_json
danielhanchen
committed
74acc40

Navigation Menu

Uh oh!

Carry ggml-org#24423 (DiffusionGemma) merged onto b10630 - #107

Carry ggml-org#24423 (DiffusionGemma) merged onto b10630#107
danielhanchen wants to merge 25 commits into
diffusiongemma-24423-upstream-basefrom
diffusiongemma-24423-b10630

Commits on Jun 10, 2026

diffusion-visual updates

diffusion: fix Windows build, skip diffusion-gemma in test-llama-archs, drop debug hooks

Commits on Jun 11, 2026

diffusion-cli: note that --fit is not applied by the diffusion runner

diffusion-cli: honor -ot / --n-cpu-moe (copy tensor_buft_overrides into model params)

diffusion: device-resident self-conditioning + cli throughput, /help, /clear

diffusion: device-side sampling reductions (default on, ~1.4x/step)

ggml-cuda: fix diffusion sampler build on HIP/MUSA

diffusion: add gemma visual server for streaming canvas frames

diffusion: self-tokenize in gemma visual server

diffusion: stop per-step device-sample fallback spam

Commits on Jun 12, 2026

diffusion-gemma-visual-server: emit STATS, auto-size MAXTOK, report toolong budget

diffusion-gemma-visual-server: split decode time from visualization overhead in STATS

Load dynamic ggml backends so the diffusion-gemma servers offload to GPU

diffusion: drop the DG_SC_CHECK / DG_DEVSAMPLE_CHECK debug verification

Commits on Jun 13, 2026

diffusion-gemma-visual-server: emit channel markers + low-VRAM auto-size fallback

Commits on Jun 14, 2026

diffusion-gemma-visual-server: size context by RAM when the model spills off the GPU

Commits on Jun 17, 2026

diffusion-gemma: chunked causal prefill + cap encode outputs to the canvas

Commits on Jun 23, 2026

Merge remote-tracking branch 'origin/master' into diffusion-visual-updates

diffusion-gemma: enable device sampler on ROCm and MUSA backends

Commits on Jul 3, 2026

Fix merge conflicts

Commits on Jul 4, 2026

Add tool calling

Commits on Aug 10, 2026

Fix merge conflicts

Fix merge conflicts

Commits on Aug 26, 2026

Merge b10630 into the DiffusionGemma pin

diffusion-gemma : port the visual server to common_json