Skip to content

ggml webgpu: actually add softmax, fix rms_norm offset #16400

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Oct 5, 2025
+10 −1
Merged

ggml webgpu: actually add softmax, fix rms_norm offset #16400

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6464269
implement soft_max
reeselevine Sep 24, 2025
ee47adb
Fix soft_max data race
reeselevine Oct 3, 2025
84770b4
Temporary fix, wait on each submit
reeselevine Oct 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 8 additions & 0 deletions ggml/src/ggml-webgpu/ggml-webgpu.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -424,6 +424,7 @@ static void ggml_backend_webgpu_build_and_enqueue(webgpu_context &
ctx->staged_param_bufs.push_back(params_bufs);
if (ctx->staged_command_bufs.size() == WEBGPU_COMMAND_SUBMIT_BATCH_SIZE) {
ggml_backend_webgpu_submit_queue(ctx);
ggml_backend_webgpu_wait_on_submission(ctx);
}
}
}
Expand Down Expand Up @@ -1060,6 +1061,9 @@ static bool ggml_webgpu_encode_node(webgpu_context ctx, ggml_tensor * node) {
case GGML_OP_SCALE:
ggml_webgpu_scale(ctx, src0, node);
break;
case GGML_OP_SOFT_MAX:
ggml_webgpu_soft_max(ctx, src0, src1, src2, node);
break;
default:
return false;
}
Expand Down Expand Up @@ -1806,6 +1810,9 @@ static bool ggml_backend_webgpu_device_supports_op(ggml_backend_dev_t dev, const
case GGML_OP_SCALE:
supports_op = op->type == GGML_TYPE_F32;
break;
case GGML_OP_SOFT_MAX:
supports_op = op->type == GGML_TYPE_F32;
break;
default:
break;
}
Expand Down Expand Up @@ -1949,6 +1956,7 @@ static ggml_backend_dev_t ggml_backend_webgpu_reg_get_device(ggml_backend_reg_t
ggml_webgpu_init_rope_pipeline(ctx);
ggml_webgpu_init_glu_pipeline(ctx);
ggml_webgpu_init_scale_pipeline(ctx);
ggml_webgpu_init_soft_max_pipeline(ctx);

#ifdef GGML_WEBGPU_DEBUG
// Initialize debug buffers
Expand Down
2 changes: 1 addition & 1 deletion ggml/src/ggml-webgpu/wgsl-shaders/rms_norm.wgsl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -84,7 +84,7 @@ fn main(@builtin(workgroup_id) wid: vec3<u32>,
let i2 = i / params.ne1;
let i1 = i % params.ne1;
let i_src_row = params.offset_src + i3 * params.stride_src3 + i2 * params.stride_src2 + i1 * params.stride_src1;
let i_dst_row = params.offset_src + i3 * params.stride_dst3 + i2 * params.stride_dst2 + i1 * params.stride_dst1;
let i_dst_row = params.offset_dst + i3 * params.stride_dst3 + i2 * params.stride_dst2 + i1 * params.stride_dst1;

let elems = (params.ne0 + wg_size - 1) / wg_size;

Expand Down
1 change: 1 addition & 0 deletions ggml/src/ggml-webgpu/wgsl-shaders/soft_max.tmpl.wgsl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -300,6 +300,7 @@ fn main(@builtin(workgroup_id) wid: vec3<u32>,
workgroupBarrier();
}
let row_max = scratch[0];
workgroupBarrier();

var sum = 0.0f;
col = lid.x;
Expand Down
Loading