Add qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5 (attention bounded by the used context) (part 3)
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/0f22bffe633c43ee9cc9d88dd97725a7b0171c08b4c0e3f7fe3a53ad985f6cf4.wgsl +29 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/19d5b2f3b725edaeae396debb3683dda307ae239c5342e1b079a2e01328b6075.wgsl +29 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/1be156f874ee05b766376dcddd23bbce11a1feb3c5dae1bd7d35869c8da15763.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/1d0e64f42c129b2019e2b9c0bd7ae7718b3b8bd67afa0dd6ae462619cc6401d1.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/237127aa3787a2a94feb4244d1b1efe7c3e5c8b5cd9e0ee544a0e99638215ac9.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/2bc311be08e7b4d612eda975fc9d13df28242058c479729c041a68fda6404f16.wgsl +12 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/4185ea53905ca7c884728b8f221f459d17104de04ad9cadbad887a4c9f75cc70.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/43ea6958ff4adf11ea8ba706a86d3b1a2a3f54d2b94ff5e795680fc33d552892.wgsl +22 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/49be6278b2b5df6dd69f6ff38e1e73dcfebc38d79f1f2694d34101dcb5c7476f.wgsl +29 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/55202c927e8c2e4f319a6e89a594e7524c7c36d903219eccd579c5b9b53ebf70.wgsl +17 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/558739fe02ea790140335a119fbd59441cc9b64320b5bfdff4daee09d3e883e0.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5814d60ecf77c9336c3118f792e0bf32f3b08128ae7987f7e535598864c1a6a1.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/587dd6b2a30d3945c4c8a5e00d7e8fa4d00fe491838abba0dd670f56af7ddb1f.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5ab3a53d6787b4ebe7eb4dae3c2da6962874d0c8cb5f23117d03473c6c76a209.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5b3a06020ba272e83296e7f824f0f46ec379e0ceb428dda7760cccea4fda9cb2.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5ca8bfbf9adfc79a462d5daf66dcee8c547b376772b0edcd07f9a084f90a17d2.wgsl +16 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/6619fc22f4476c11499f21aed05ce1963328fe028d204aa287d1414d3222b564.wgsl +34 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/67991f0575b151ad2e89edf7eca755a0d62686cddce367e5d3d5382518167586.wgsl +12 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/6b83f8154d25e77d442948180ca249a492f7d9490a41d3dda46cd9bcb2fb37fd.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/707b7b188f55bd9c37b6bad7708d7b38bc286b7147e6355716910e7a44f9ae82.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/80b140b8a7c9a19ca722fbb3b863a3d3640c28ffb7466d84f26433a8e5cbcf17.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/83993849c704ba6fb220c0b8a0fd6a3847d253423143816289ff7e939dd204f8.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/855590b18d4dbaabbe80b309cf365dfc1beb2e6ee716d5022c5d4c855287c7c3.wgsl +11 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/88a9af67da30bf6c4a9f13401f1a21aa87f0731e780bbf00d45e23bec4aa5591.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/88e625463de0ca80326d53aa24f84a402f18e2e96f9ec322c6a1ac2dea47cfbf.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/9070ec757115bd7c81186acd9d29ce7bc0258c77cbd02866b5dfb9c54eecd035.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/984f5c91f469bb3fdfb0d8ddfe082a64ba3dfe87fc9dc4c1db582ca1f62d5f2b.wgsl +29 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/9d29d01524ce6f64fd00438ed7f03ce6796d0f22be1be262d6fbe43c51ffd2b6.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ab1e7db44b714b30926b7a0f1c35175c6b3477b2cc9c2e8bd5f025cff4c61200.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/bf3ebcda6a36c384622b5a3d36a8c118cb368061ec695b8bc0fb46b550faf09b.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/c09ba2ff8e99d3ef1fa686820ba2f6b198f2bd5f7755c8ee15dbf73344b641e8.wgsl +17 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/c667300f7d1060bdf36ea2a09ce4905b84ff7e9d1711fcc29c947425e0126361.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/cbb27a57c814432e940e67eb4f69d697bcb0099119a6bc46d85ed252de5cd9c1.wgsl +29 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/cd6a942d0e0231d1e8fd05a2ce362512dfaed46ceba46fc86593182de4a7a076.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ce90f1e7d902eecec3952a34ca17617870d27474379f6e7c82faee10f8376e5a.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/d16842c527e3b75143d68b32c2099a38e0195b20ce1b8664132a79bad3bcff4d.wgsl +18 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/dcc64ada7c202b3c7a4fa83dbec165a67733352558b015eff6047e5e250907af.wgsl +11 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ea188fbe79cee6ca9210d873c95f40fd3edd4c7140e6b2eb4739fedb2678e232.wgsl +17 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ee196d65aaf47e2ddd77b2315e4c555b915f83d661650ce4953975023efbabd6.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/f92c5546622dc3aea6badf14abe019d854d3d9cb34a315d841bac05f9eb91066.wgsl +10 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/fbf9f7c699b99624e8563e1f8fab5b68cf89160d60e3daf5945f1ffe5a8954c3.wgsl +9 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/manifest.json +316 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/source/LICENSE +202 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/source/config.json +96 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00000.bin +3 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00001.bin +3 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00002.bin +3 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00003.bin +3 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00004.bin +3 -0
- qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00005.bin +3 -0
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/0f22bffe633c43ee9cc9d88dd97725a7b0171c08b4c0e3f7fe3a53ad985f6cf4.wgsl
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 6 |
+
let pair = b1[index / 2u];
|
| 7 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 8 |
+
return bitcast<f32>(bits << 16u);
|
| 9 |
+
}
|
| 10 |
+
|
| 11 |
+
var<workgroup> tile_a: array<f32, 64>;
|
| 12 |
+
var<workgroup> tile_b: array<f32, 64>;
|
| 13 |
+
@compute @workgroup_size(8, 8)
|
| 14 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 15 |
+
let row = group.y * 8u + local.y;
|
| 16 |
+
let col = group.x * 8u + local.x;
|
| 17 |
+
let i = (group.z * 247u + row) * 2048u + col;
|
| 18 |
+
var acc = 0.0;
|
| 19 |
+
for (var tile = 0u; tile < 4096u; tile += 8u) {
|
| 20 |
+
tile_a[local.y * 8u + local.x] = 0.0;
|
| 21 |
+
tile_b[local.x * 8u + local.y] = 0.0;
|
| 22 |
+
if (row < 247u && tile + local.x < 4096u) { tile_a[local.y * 8u + local.x] = f32(b0[(group.z * 247u + row) * 4096u + tile + local.x]); }
|
| 23 |
+
if (group.x * 8u + local.y < 2048u && tile + local.x < 4096u) { tile_b[local.x * 8u + local.y] = unpack_bf16_1((group.x * 8u + local.y) * 4096u + tile + local.x); }
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var p = 0u; p < 8u; p++) { acc += tile_a[local.y * 8u + p] * tile_b[p * 8u + local.x]; }
|
| 26 |
+
workgroupBarrier();
|
| 27 |
+
}
|
| 28 |
+
if (row < 247u && col < 2048u) { out[i] = f32(acc + f32(b2[col])); }
|
| 29 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/19d5b2f3b725edaeae396debb3683dda307ae239c5342e1b079a2e01328b6075.wgsl
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 6 |
+
let pair = b1[index / 2u];
|
| 7 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 8 |
+
return bitcast<f32>(bits << 16u);
|
| 9 |
+
}
|
| 10 |
+
|
| 11 |
+
var<workgroup> tile_a: array<f32, 64>;
|
| 12 |
+
var<workgroup> tile_b: array<f32, 64>;
|
| 13 |
+
@compute @workgroup_size(8, 8)
|
| 14 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 15 |
+
let row = group.y * 8u + local.y;
|
| 16 |
+
let col = group.x * 8u + local.x;
|
| 17 |
+
let i = (group.z * 988u + row) * 3072u + col;
|
| 18 |
+
var acc = 0.0;
|
| 19 |
+
for (var tile = 0u; tile < 1024u; tile += 8u) {
|
| 20 |
+
tile_a[local.y * 8u + local.x] = 0.0;
|
| 21 |
+
tile_b[local.x * 8u + local.y] = 0.0;
|
| 22 |
+
if (row < 988u && tile + local.x < 1024u) { tile_a[local.y * 8u + local.x] = f32(b0[(group.z * 988u + row) * 1024u + tile + local.x]); }
|
| 23 |
+
if (group.x * 8u + local.y < 3072u && tile + local.x < 1024u) { tile_b[local.x * 8u + local.y] = unpack_bf16_1((group.x * 8u + local.y) * 1024u + tile + local.x); }
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var p = 0u; p < 8u; p++) { acc += tile_a[local.y * 8u + p] * tile_b[p * 8u + local.x]; }
|
| 26 |
+
workgroupBarrier();
|
| 27 |
+
}
|
| 28 |
+
if (row < 988u && col < 3072u) { out[i] = f32(acc + f32(b2[col])); }
|
| 29 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/1be156f874ee05b766376dcddd23bbce11a1feb3c5dae1bd7d35869c8da15763.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[1u * 1011712u + ((i / 1024u) % 988u) * 1024u + ((i / 64u) % 16u) * 64u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/1d0e64f42c129b2019e2b9c0bd7ae7718b3b8bd67afa0dd6ae462619cc6401d1.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 1011712u) % 1u) * 1011712u + ((i / 63232u) % 16u) * 63232u + ((i / 988u) % 64u) * 1u + ((i / 1u) % 988u) * 64u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/237127aa3787a2a94feb4244d1b1efe7c3e5c8b5cd9e0ee544a0e99638215ac9.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<i32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<i32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1984u;
|
| 7 |
+
if (i >= 1976u) { return; }
|
| 8 |
+
out[i] = i32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/2bc311be08e7b4d612eda975fc9d13df28242058c479729c041a68fda6404f16.wgsl
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 1011712u;
|
| 8 |
+
if (i >= 1011712u) { return; }
|
| 9 |
+
let coord = (i / 1u) % 64u;
|
| 10 |
+
if (coord >= 0u && coord < 32u) { out[i] = f32(b0[(i / 64u) * 32u + (coord - 0u) * 1u + i % 1u]); }
|
| 11 |
+
if (coord >= 32u && coord < 64u) { out[i] = f32(b1[(i / 64u) * 32u + (coord - 32u) * 1u + i % 1u]); }
|
| 12 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/4185ea53905ca7c884728b8f221f459d17104de04ad9cadbad887a4c9f75cc70.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 63232u;
|
| 7 |
+
if (i >= 63232u) { return; }
|
| 8 |
+
let x = f32(b0[i]);
|
| 9 |
+
out[i] = f32(sin(x));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/43ea6958ff4adf11ea8ba706a86d3b1a2a3f54d2b94ff5e795680fc33d552892.wgsl
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
|
| 6 |
+
@compute @workgroup_size(64)
|
| 7 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 8 |
+
let i = gid.x + gid.y * 1011712u;
|
| 9 |
+
if (i >= 1011712u) { return; }
|
| 10 |
+
let batch = i / 1024u;
|
| 11 |
+
let channel = (i / 1u) % 1024u;
|
| 12 |
+
let group = channel / 1024u;
|
| 13 |
+
var acc = 0.0;
|
| 14 |
+
for (var c = 0u; c < 3u; c++) {
|
| 15 |
+
for (var p = 0u; p < 512u; p++) {
|
| 16 |
+
let pos0 = i32((i / 1u) % 1u) * 2 + i32((p / 256u) % 2u) * 1 - 0; let pos1 = i32((i / 1u) % 1u) * 16 + i32((p / 16u) % 16u) * 1 - 0; let pos2 = i32((i / 1u) % 1u) * 16 + i32((p / 1u) % 16u) * 1 - 0;
|
| 17 |
+
if (pos0 >= 0 && pos0 < 2 && pos1 >= 0 && pos1 < 16 && pos2 >= 0 && pos2 < 16) { acc += f32(b1[batch * 1536u + (group * 3u + c) * 512u + u32(pos0) * 256u + u32(pos1) * 16u + u32(pos2) * 1u]) * f32(b2[(channel * 3u + c) * 512u + p]); }
|
| 18 |
+
}
|
| 19 |
+
}
|
| 20 |
+
out[i] = f32(acc + f32(b0[channel]));
|
| 21 |
+
|
| 22 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/49be6278b2b5df6dd69f6ff38e1e73dcfebc38d79f1f2694d34101dcb5c7476f.wgsl
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 6 |
+
let pair = b1[index / 2u];
|
| 7 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 8 |
+
return bitcast<f32>(bits << 16u);
|
| 9 |
+
}
|
| 10 |
+
|
| 11 |
+
var<workgroup> tile_a: array<f32, 64>;
|
| 12 |
+
var<workgroup> tile_b: array<f32, 64>;
|
| 13 |
+
@compute @workgroup_size(8, 8)
|
| 14 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 15 |
+
let row = group.y * 8u + local.y;
|
| 16 |
+
let col = group.x * 8u + local.x;
|
| 17 |
+
let i = (group.z * 988u + row) * 1024u + col;
|
| 18 |
+
var acc = 0.0;
|
| 19 |
+
for (var tile = 0u; tile < 4096u; tile += 8u) {
|
| 20 |
+
tile_a[local.y * 8u + local.x] = 0.0;
|
| 21 |
+
tile_b[local.x * 8u + local.y] = 0.0;
|
| 22 |
+
if (row < 988u && tile + local.x < 4096u) { tile_a[local.y * 8u + local.x] = f32(b0[(group.z * 988u + row) * 4096u + tile + local.x]); }
|
| 23 |
+
if (group.x * 8u + local.y < 1024u && tile + local.x < 4096u) { tile_b[local.x * 8u + local.y] = unpack_bf16_1((group.x * 8u + local.y) * 4096u + tile + local.x); }
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var p = 0u; p < 8u; p++) { acc += tile_a[local.y * 8u + p] * tile_b[p * 8u + local.x]; }
|
| 26 |
+
workgroupBarrier();
|
| 27 |
+
}
|
| 28 |
+
if (row < 988u && col < 1024u) { out[i] = f32(acc + f32(b2[col])); }
|
| 29 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/55202c927e8c2e4f319a6e89a594e7524c7c36d903219eccd579c5b9b53ebf70.wgsl
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 1011712u;
|
| 8 |
+
if (i >= 1011712u) { return; }
|
| 9 |
+
let batch = i / 63232u;
|
| 10 |
+
let row = (i / 64u) % 988u;
|
| 11 |
+
let col = i % 64u;
|
| 12 |
+
var acc = 0.0;
|
| 13 |
+
for (var p = 0u; p < 988u; p++) {
|
| 14 |
+
acc += f32(b0[(((batch / 1u) % 16u) * 976144u) + row * 988u + p]) * f32(b1[(((batch / 1u) % 16u) * 63232u) + p * 64u + col]);
|
| 15 |
+
}
|
| 16 |
+
out[i] = f32(acc);
|
| 17 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/558739fe02ea790140335a119fbd59441cc9b64320b5bfdff4daee09d3e883e0.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[2u * 1011712u + ((i / 1024u) % 988u) * 1024u + ((i / 64u) % 16u) * 64u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5814d60ecf77c9336c3118f792e0bf32f3b08128ae7987f7e535598864c1a6a1.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 1011712u) % 1u) * 1011712u + ((i / 1024u) % 988u) * 64u + ((i / 64u) % 16u) * 63232u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/587dd6b2a30d3945c4c8a5e00d7e8fa4d00fe491838abba0dd670f56af7ddb1f.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 4046848u;
|
| 8 |
+
if (i >= 4046848u) { return; }
|
| 9 |
+
out[i] = f32(f32(b0[i]) * f32(b1[((i / 4096u) % 988u) * 4u + ((i / 1024u) % 4u) * 1u]));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5ab3a53d6787b4ebe7eb4dae3c2da6962874d0c8cb5f23117d03473c6c76a209.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 505856u;
|
| 7 |
+
if (i >= 505856u) { return; }
|
| 8 |
+
let x = f32(b0[i]);
|
| 9 |
+
out[i] = f32(-x);
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5b3a06020ba272e83296e7f824f0f46ec379e0ceb428dda7760cccea4fda9cb2.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 31616u;
|
| 7 |
+
if (i >= 31616u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/5ca8bfbf9adfc79a462d5daf66dcee8c547b376772b0edcd07f9a084f90a17d2.wgsl
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<i32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 5 |
+
let pair = b1[index / 2u];
|
| 6 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 7 |
+
return bitcast<f32>(bits << 16u);
|
| 8 |
+
}
|
| 9 |
+
|
| 10 |
+
@compute @workgroup_size(64)
|
| 11 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 12 |
+
let i = gid.x + gid.y * 4046848u;
|
| 13 |
+
if (i >= 4046848u) { return; }
|
| 14 |
+
let token = i32(b0[i / 1024u]);
|
| 15 |
+
out[i] = f32(unpack_bf16_1(u32(token - 0) * 1024u + i % 1024u));
|
| 16 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/6619fc22f4476c11499f21aed05ce1963328fe028d204aa287d1414d3222b564.wgsl
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
var<workgroup> partial: array<f32, 64>;
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 7 |
+
let row = group.x + group.y * 15808u;
|
| 8 |
+
if (row >= 15808u) { return; }
|
| 9 |
+
let lane = local.x;
|
| 10 |
+
var maximum = -3.4028234663852886e38;
|
| 11 |
+
for (var p = lane; p < 988u; p += 64u) { maximum = max(maximum, f32(b0[row * 988u + p])); }
|
| 12 |
+
partial[lane] = maximum;
|
| 13 |
+
workgroupBarrier();
|
| 14 |
+
for (var stride = 32u; stride > 0u; stride /= 2u) {
|
| 15 |
+
if (lane < stride) { partial[lane] = max(partial[lane], partial[lane + stride]); }
|
| 16 |
+
workgroupBarrier();
|
| 17 |
+
}
|
| 18 |
+
maximum = partial[0];
|
| 19 |
+
// All lanes must finish reading the maximum before the scratch array is reused.
|
| 20 |
+
workgroupBarrier();
|
| 21 |
+
var total = 0.0;
|
| 22 |
+
for (var p = lane; p < 988u; p += 64u) { total += exp(f32(b0[row * 988u + p]) - maximum); }
|
| 23 |
+
partial[lane] = total;
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var stride = 32u; stride > 0u; stride /= 2u) {
|
| 26 |
+
if (lane < stride) { partial[lane] += partial[lane + stride]; }
|
| 27 |
+
workgroupBarrier();
|
| 28 |
+
}
|
| 29 |
+
total = partial[0];
|
| 30 |
+
for (var p = lane; p < 988u; p += 64u) {
|
| 31 |
+
let i = row * 988u + p;
|
| 32 |
+
out[i] = f32(exp(f32(b0[row * 988u + p]) - maximum) / total);
|
| 33 |
+
}
|
| 34 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/67991f0575b151ad2e89edf7eca755a0d62686cddce367e5d3d5382518167586.wgsl
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 4046848u;
|
| 7 |
+
if (i >= 4046848u) { return; }
|
| 8 |
+
let x = f32(b0[i]);
|
| 9 |
+
if (x > 10.0) { out[i] = f32(x); return; }
|
| 10 |
+
if (x < -10.0) { out[i] = f32(0.0); return; }
|
| 11 |
+
out[i] = f32(0.5 * x * (1.0 + tanh(clamp(0.7978845608028654 * (x + 0.044715 * x*x*x), -10.0, 10.0))));
|
| 12 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/6b83f8154d25e77d442948180ca249a492f7d9490a41d3dda46cd9bcb2fb37fd.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 4194240u;
|
| 7 |
+
if (i >= 15618304u) { return; }
|
| 8 |
+
out[i] = f32(f32(b0[i]) * 0.125);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/707b7b188f55bd9c37b6bad7708d7b38bc286b7147e6355716910e7a44f9ae82.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 3035136u;
|
| 7 |
+
if (i >= 3035136u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/80b140b8a7c9a19ca722fbb3b863a3d3640c28ffb7466d84f26433a8e5cbcf17.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 63232u;
|
| 7 |
+
if (i >= 63232u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/83993849c704ba6fb220c0b8a0fd6a3847d253423143816289ff7e939dd204f8.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
let coord = (i / 1024u) % 988u;
|
| 9 |
+
if (coord >= 0u && coord < 988u) { out[i] = f32(b0[(i / 1011712u) * 1011712u + (coord - 0u) * 1024u + i % 1024u]); }
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/855590b18d4dbaabbe80b309cf365dfc1beb2e6ee716d5022c5d4c855287c7c3.wgsl
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 63232u;
|
| 7 |
+
if (i >= 63232u) { return; }
|
| 8 |
+
let coord = (i / 1u) % 64u;
|
| 9 |
+
if (coord >= 0u && coord < 32u) { out[i] = f32(b0[(i / 64u) * 32u + (coord - 0u) * 1u + i % 1u]); }
|
| 10 |
+
if (coord >= 32u && coord < 64u) { out[i] = f32(b0[(i / 64u) * 32u + (coord - 32u) * 1u + i % 1u]); }
|
| 11 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/88a9af67da30bf6c4a9f13401f1a21aa87f0731e780bbf00d45e23bec4aa5591.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 3968u;
|
| 7 |
+
if (i >= 3952u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/88e625463de0ca80326d53aa24f84a402f18e2e96f9ec322c6a1ac2dea47cfbf.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 3035136u;
|
| 7 |
+
if (i >= 3035136u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 1011712u) % 3u) * 1024u + ((i / 1024u) % 988u) * 3072u + ((i / 64u) % 16u) * 64u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/9070ec757115bd7c81186acd9d29ce7bc0258c77cbd02866b5dfb9c54eecd035.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 505856u;
|
| 7 |
+
if (i >= 505856u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 512u) % 988u) * 1024u + ((i / 32u) % 16u) * 64u + (((i / 1u) % 32u) * 1u + 0u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/984f5c91f469bb3fdfb0d8ddfe082a64ba3dfe87fc9dc4c1db582ca1f62d5f2b.wgsl
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 6 |
+
let pair = b1[index / 2u];
|
| 7 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 8 |
+
return bitcast<f32>(bits << 16u);
|
| 9 |
+
}
|
| 10 |
+
|
| 11 |
+
var<workgroup> tile_a: array<f32, 64>;
|
| 12 |
+
var<workgroup> tile_b: array<f32, 64>;
|
| 13 |
+
@compute @workgroup_size(8, 8)
|
| 14 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 15 |
+
let row = group.y * 8u + local.y;
|
| 16 |
+
let col = group.x * 8u + local.x;
|
| 17 |
+
let i = (group.z * 247u + row) * 4096u + col;
|
| 18 |
+
var acc = 0.0;
|
| 19 |
+
for (var tile = 0u; tile < 4096u; tile += 8u) {
|
| 20 |
+
tile_a[local.y * 8u + local.x] = 0.0;
|
| 21 |
+
tile_b[local.x * 8u + local.y] = 0.0;
|
| 22 |
+
if (row < 247u && tile + local.x < 4096u) { tile_a[local.y * 8u + local.x] = f32(b0[(group.z * 247u + row) * 4096u + tile + local.x]); }
|
| 23 |
+
if (group.x * 8u + local.y < 4096u && tile + local.x < 4096u) { tile_b[local.x * 8u + local.y] = unpack_bf16_1((group.x * 8u + local.y) * 4096u + tile + local.x); }
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var p = 0u; p < 8u; p++) { acc += tile_a[local.y * 8u + p] * tile_b[p * 8u + local.x]; }
|
| 26 |
+
workgroupBarrier();
|
| 27 |
+
}
|
| 28 |
+
if (row < 247u && col < 4096u) { out[i] = f32(acc + f32(b2[col])); }
|
| 29 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/9d29d01524ce6f64fd00438ed7f03ce6796d0f22be1be262d6fbe43c51ffd2b6.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 63232u) % 16u) * 64u + ((i / 64u) % 988u) * 1024u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ab1e7db44b714b30926b7a0f1c35175c6b3477b2cc9c2e8bd5f025cff4c61200.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 1011712u;
|
| 8 |
+
if (i >= 1011712u) { return; }
|
| 9 |
+
out[i] = f32(f32(b0[i]) + (f32(b1[i]) * 1.0));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/bf3ebcda6a36c384622b5a3d36a8c118cb368061ec695b8bc0fb46b550faf09b.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/c09ba2ff8e99d3ef1fa686820ba2f6b198f2bd5f7755c8ee15dbf73344b641e8.wgsl
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 4194240u;
|
| 8 |
+
if (i >= 15618304u) { return; }
|
| 9 |
+
let batch = i / 976144u;
|
| 10 |
+
let row = (i / 988u) % 988u;
|
| 11 |
+
let col = i % 988u;
|
| 12 |
+
var acc = 0.0;
|
| 13 |
+
for (var p = 0u; p < 64u; p++) {
|
| 14 |
+
acc += f32(b0[(((batch / 1u) % 16u) * 63232u) + row * 64u + p]) * f32(b1[(((batch / 1u) % 16u) * 63232u) + p * 988u + col]);
|
| 15 |
+
}
|
| 16 |
+
out[i] = f32(acc);
|
| 17 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/c667300f7d1060bdf36ea2a09ce4905b84ff7e9d1711fcc29c947425e0126361.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[((i / 1011712u) % 1u) * 1011712u + ((i / 63232u) % 16u) * 63232u + (((i / 64u) % 988u) * 1u + 0u) * 64u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/cbb27a57c814432e940e67eb4f69d697bcb0099119a6bc46d85ed252de5cd9c1.wgsl
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<u32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
fn unpack_bf16_1(index: u32) -> f32 {
|
| 6 |
+
let pair = b1[index / 2u];
|
| 7 |
+
let bits = (pair >> ((index % 2u) * 16u)) & 65535u;
|
| 8 |
+
return bitcast<f32>(bits << 16u);
|
| 9 |
+
}
|
| 10 |
+
|
| 11 |
+
var<workgroup> tile_a: array<f32, 64>;
|
| 12 |
+
var<workgroup> tile_b: array<f32, 64>;
|
| 13 |
+
@compute @workgroup_size(8, 8)
|
| 14 |
+
fn main(@builtin(workgroup_id) group: vec3<u32>, @builtin(local_invocation_id) local: vec3<u32>) {
|
| 15 |
+
let row = group.y * 8u + local.y;
|
| 16 |
+
let col = group.x * 8u + local.x;
|
| 17 |
+
let i = (group.z * 988u + row) * 1024u + col;
|
| 18 |
+
var acc = 0.0;
|
| 19 |
+
for (var tile = 0u; tile < 1024u; tile += 8u) {
|
| 20 |
+
tile_a[local.y * 8u + local.x] = 0.0;
|
| 21 |
+
tile_b[local.x * 8u + local.y] = 0.0;
|
| 22 |
+
if (row < 988u && tile + local.x < 1024u) { tile_a[local.y * 8u + local.x] = f32(b0[(group.z * 988u + row) * 1024u + tile + local.x]); }
|
| 23 |
+
if (group.x * 8u + local.y < 1024u && tile + local.x < 1024u) { tile_b[local.x * 8u + local.y] = unpack_bf16_1((group.x * 8u + local.y) * 1024u + tile + local.x); }
|
| 24 |
+
workgroupBarrier();
|
| 25 |
+
for (var p = 0u; p < 8u; p++) { acc += tile_a[local.y * 8u + p] * tile_b[p * 8u + local.x]; }
|
| 26 |
+
workgroupBarrier();
|
| 27 |
+
}
|
| 28 |
+
if (row < 988u && col < 1024u) { out[i] = f32(acc + f32(b2[col])); }
|
| 29 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/cd6a942d0e0231d1e8fd05a2ce362512dfaed46ceba46fc86593182de4a7a076.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
out[i] = f32(b0[0u * 1011712u + ((i / 1024u) % 988u) * 1024u + ((i / 64u) % 16u) * 64u + ((i / 1u) % 64u) * 1u]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ce90f1e7d902eecec3952a34ca17617870d27474379f6e7c82faee10f8376e5a.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<i32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 31616u;
|
| 8 |
+
if (i >= 31616u) { return; }
|
| 9 |
+
out[i] = f32(f32(b0[((i / 32u) % 988u) * 2u + ((i / 16u) % 2u) * 1u]) * f32(b1[((i / 1u) % 16u) * 1u]));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/d16842c527e3b75143d68b32c2099a38e0195b20ce1b8664132a79bad3bcff4d.wgsl
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read> b2: array<f32>;
|
| 4 |
+
@group(0) @binding(3) var<storage, read_write> out: array<f32>;
|
| 5 |
+
|
| 6 |
+
@compute @workgroup_size(64)
|
| 7 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 8 |
+
let i = gid.x + gid.y * 1011712u;
|
| 9 |
+
if (i >= 1011712u) { return; }
|
| 10 |
+
let base = (i / 1024u) * 1024u;
|
| 11 |
+
var mean = 0.0;
|
| 12 |
+
for (var j = 0u; j < 1024u; j++) { mean += f32(b0[base + j]); }
|
| 13 |
+
mean /= 1024.0;
|
| 14 |
+
var variance = 0.0;
|
| 15 |
+
for (var j = 0u; j < 1024u; j++) { let centered = f32(b0[base + j]) - mean; variance += centered * centered; }
|
| 16 |
+
out[i] = f32((f32(b0[i]) - mean) * inverseSqrt(variance / 1024.0 + 1e-06) * f32(b1[i % 1024u]) + f32(b2[i % 1024u]));
|
| 17 |
+
|
| 18 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/dcc64ada7c202b3c7a4fa83dbec165a67733352558b015eff6047e5e250907af.wgsl
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
var acc = 0.0;
|
| 9 |
+
for (var j = 0u; j < 4u; j++) { acc += f32(b0[(((i / 1024u) % 988u) * 4096u + ((i / 1u) % 1024u) * 1u) + (((j / 1u) % 4u) * 1024u)]); }
|
| 10 |
+
out[i] = f32(acc);
|
| 11 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ea188fbe79cee6ca9210d873c95f40fd3edd4c7140e6b2eb4739fedb2678e232.wgsl
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1011712u;
|
| 7 |
+
if (i >= 1011712u) { return; }
|
| 8 |
+
let x = f32(b0[i]);
|
| 9 |
+
if (x > 10.0) { out[i] = f32(x); return; }
|
| 10 |
+
if (x < -10.0) { out[i] = f32(0.0); return; }
|
| 11 |
+
let z = x * 0.7071067811865476;
|
| 12 |
+
let t = 1.0 / (1.0 + 0.3275911 * abs(z));
|
| 13 |
+
let polynomial = (((((1.061405429 * t - 1.453152027) * t) + 1.421413741) * t - 0.284496736) * t + 0.254829592) * t;
|
| 14 |
+
let erf = sign(z) * (1.0 - polynomial * exp(-z * z));
|
| 15 |
+
out[i] = f32(0.5 * x * (1.0 + erf));
|
| 16 |
+
|
| 17 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/ee196d65aaf47e2ddd77b2315e4c555b915f83d661650ce4953975023efbabd6.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 63232u;
|
| 7 |
+
if (i >= 63232u) { return; }
|
| 8 |
+
let x = f32(b0[i]);
|
| 9 |
+
out[i] = f32(cos(x));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/f92c5546622dc3aea6badf14abe019d854d3d9cb34a315d841bac05f9eb91066.wgsl
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read> b1: array<f32>;
|
| 3 |
+
@group(0) @binding(2) var<storage, read_write> out: array<f32>;
|
| 4 |
+
|
| 5 |
+
@compute @workgroup_size(64)
|
| 6 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 7 |
+
let i = gid.x + gid.y * 1011712u;
|
| 8 |
+
if (i >= 1011712u) { return; }
|
| 9 |
+
out[i] = f32(f32(b0[i]) * f32(b1[((i / 1024u) % 988u) * 64u + ((i / 1u) % 64u) * 1u]));
|
| 10 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/kernels/fbf9f7c699b99624e8563e1f8fab5b68cf89160d60e3daf5945f1ffe5a8954c3.wgsl
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
@group(0) @binding(0) var<storage, read> b0: array<f32>;
|
| 2 |
+
@group(0) @binding(1) var<storage, read_write> out: array<f32>;
|
| 3 |
+
|
| 4 |
+
@compute @workgroup_size(64)
|
| 5 |
+
fn main(@builtin(global_invocation_id) gid: vec3<u32>) {
|
| 6 |
+
let i = gid.x + gid.y * 1517568u;
|
| 7 |
+
if (i >= 1517568u) { return; }
|
| 8 |
+
out[i] = f32(b0[i]);
|
| 9 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/manifest.json
ADDED
|
@@ -0,0 +1,316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"format": "webtorch",
|
| 3 |
+
"version": 1,
|
| 4 |
+
"producer": {
|
| 5 |
+
"name": "webtorch",
|
| 6 |
+
"version": "0.1.0a1",
|
| 7 |
+
"torch": "2.14.0"
|
| 8 |
+
},
|
| 9 |
+
"entrypoints": {
|
| 10 |
+
"forward": "graph.json"
|
| 11 |
+
},
|
| 12 |
+
"requiredFeatures": [],
|
| 13 |
+
"assets": [
|
| 14 |
+
{
|
| 15 |
+
"path": "weights/00000.bin",
|
| 16 |
+
"byteLength": 67108864,
|
| 17 |
+
"sha256": "5ff01e44eac7d7a30d3779fe7970f1ec882ae3a807ded59b8c25b203b57bd746"
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"path": "weights/00001.bin",
|
| 21 |
+
"byteLength": 67108864,
|
| 22 |
+
"sha256": "3f698d7140ff322989ac4a6aa5c05976740689c7129be1f179362c94bef2730f"
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"path": "weights/00002.bin",
|
| 26 |
+
"byteLength": 67108864,
|
| 27 |
+
"sha256": "1a05e5a106c236dedfca3a681806500e8d9f518767ec432ffa7a53a0061e94f8"
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"path": "weights/00003.bin",
|
| 31 |
+
"byteLength": 67108864,
|
| 32 |
+
"sha256": "da533794b3c088224ed91b19ce038c0b07acbfa7b01eeb5b6b1e0d8bc39e5bc8"
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"path": "weights/00004.bin",
|
| 36 |
+
"byteLength": 67108864,
|
| 37 |
+
"sha256": "b95eb7e977231963dd5335fab05dc5d111c3c5c716bc0bac3e0f48af84b5cf18"
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"path": "weights/00005.bin",
|
| 41 |
+
"byteLength": 67108864,
|
| 42 |
+
"sha256": "17c20eaa5be55412038d213e9820d01a4f1f822678a1fbe5d7df20b5012b546e"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"path": "weights/00006.bin",
|
| 46 |
+
"byteLength": 67108864,
|
| 47 |
+
"sha256": "65031b11b25789c6ccbcddf46f2784210191bc64a6eca510ba6eb479f4dab7c9"
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"path": "weights/00007.bin",
|
| 51 |
+
"byteLength": 67108864,
|
| 52 |
+
"sha256": "65e01ad7b5d868aa12a669a3ac8ab2b7c5434fd3e484da3331a80608d42f3ff2"
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"path": "weights/00008.bin",
|
| 56 |
+
"byteLength": 67108864,
|
| 57 |
+
"sha256": "d1edbabeb49f1b143f5c09814a3185c0557fbeaf313d214876198ddc11ac0b6a"
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"path": "weights/00009.bin",
|
| 61 |
+
"byteLength": 62696096,
|
| 62 |
+
"sha256": "81e70e09fa4e106a132f634c3234dd50f42093f90cff24efe8732daa4a99a2fd"
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"path": "kernels/fbf9f7c699b99624e8563e1f8fab5b68cf89160d60e3daf5945f1ffe5a8954c3.wgsl",
|
| 66 |
+
"byteLength": 296,
|
| 67 |
+
"sha256": "fbf9f7c699b99624e8563e1f8fab5b68cf89160d60e3daf5945f1ffe5a8954c3"
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"path": "kernels/43ea6958ff4adf11ea8ba706a86d3b1a2a3f54d2b94ff5e795680fc33d552892.wgsl",
|
| 71 |
+
"byteLength": 1055,
|
| 72 |
+
"sha256": "43ea6958ff4adf11ea8ba706a86d3b1a2a3f54d2b94ff5e795680fc33d552892"
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"path": "kernels/bf3ebcda6a36c384622b5a3d36a8c118cb368061ec695b8bc0fb46b550faf09b.wgsl",
|
| 76 |
+
"byteLength": 296,
|
| 77 |
+
"sha256": "bf3ebcda6a36c384622b5a3d36a8c118cb368061ec695b8bc0fb46b550faf09b"
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"path": "kernels/5ca8bfbf9adfc79a462d5daf66dcee8c547b376772b0edcd07f9a084f90a17d2.wgsl",
|
| 81 |
+
"byteLength": 594,
|
| 82 |
+
"sha256": "5ca8bfbf9adfc79a462d5daf66dcee8c547b376772b0edcd07f9a084f90a17d2"
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"path": "kernels/88a9af67da30bf6c4a9f13401f1a21aa87f0731e780bbf00d45e23bec4aa5591.wgsl",
|
| 86 |
+
"byteLength": 290,
|
| 87 |
+
"sha256": "88a9af67da30bf6c4a9f13401f1a21aa87f0731e780bbf00d45e23bec4aa5591"
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"path": "kernels/587dd6b2a30d3945c4c8a5e00d7e8fa4d00fe491838abba0dd670f56af7ddb1f.wgsl",
|
| 91 |
+
"byteLength": 421,
|
| 92 |
+
"sha256": "587dd6b2a30d3945c4c8a5e00d7e8fa4d00fe491838abba0dd670f56af7ddb1f"
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"path": "kernels/dcc64ada7c202b3c7a4fa83dbec165a67733352558b015eff6047e5e250907af.wgsl",
|
| 96 |
+
"byteLength": 445,
|
| 97 |
+
"sha256": "dcc64ada7c202b3c7a4fa83dbec165a67733352558b015eff6047e5e250907af"
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"path": "kernels/ab1e7db44b714b30926b7a0f1c35175c6b3477b2cc9c2e8bd5f025cff4c61200.wgsl",
|
| 101 |
+
"byteLength": 379,
|
| 102 |
+
"sha256": "ab1e7db44b714b30926b7a0f1c35175c6b3477b2cc9c2e8bd5f025cff4c61200"
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"path": "kernels/237127aa3787a2a94feb4244d1b1efe7c3e5c8b5cd9e0ee544a0e99638215ac9.wgsl",
|
| 106 |
+
"byteLength": 290,
|
| 107 |
+
"sha256": "237127aa3787a2a94feb4244d1b1efe7c3e5c8b5cd9e0ee544a0e99638215ac9"
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"path": "kernels/ce90f1e7d902eecec3952a34ca17617870d27474379f6e7c82faee10f8376e5a.wgsl",
|
| 111 |
+
"byteLength": 433,
|
| 112 |
+
"sha256": "ce90f1e7d902eecec3952a34ca17617870d27474379f6e7c82faee10f8376e5a"
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"path": "kernels/5b3a06020ba272e83296e7f824f0f46ec379e0ceb428dda7760cccea4fda9cb2.wgsl",
|
| 116 |
+
"byteLength": 292,
|
| 117 |
+
"sha256": "5b3a06020ba272e83296e7f824f0f46ec379e0ceb428dda7760cccea4fda9cb2"
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"path": "kernels/855590b18d4dbaabbe80b309cf365dfc1beb2e6ee716d5022c5d4c855287c7c3.wgsl",
|
| 121 |
+
"byteLength": 501,
|
| 122 |
+
"sha256": "855590b18d4dbaabbe80b309cf365dfc1beb2e6ee716d5022c5d4c855287c7c3"
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"path": "kernels/ee196d65aaf47e2ddd77b2315e4c555b915f83d661650ce4953975023efbabd6.wgsl",
|
| 126 |
+
"byteLength": 313,
|
| 127 |
+
"sha256": "ee196d65aaf47e2ddd77b2315e4c555b915f83d661650ce4953975023efbabd6"
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"path": "kernels/4185ea53905ca7c884728b8f221f459d17104de04ad9cadbad887a4c9f75cc70.wgsl",
|
| 131 |
+
"byteLength": 313,
|
| 132 |
+
"sha256": "4185ea53905ca7c884728b8f221f459d17104de04ad9cadbad887a4c9f75cc70"
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"path": "kernels/d16842c527e3b75143d68b32c2099a38e0195b20ce1b8664132a79bad3bcff4d.wgsl",
|
| 136 |
+
"byteLength": 767,
|
| 137 |
+
"sha256": "d16842c527e3b75143d68b32c2099a38e0195b20ce1b8664132a79bad3bcff4d"
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"path": "kernels/19d5b2f3b725edaeae396debb3683dda307ae239c5342e1b079a2e01328b6075.wgsl",
|
| 141 |
+
"byteLength": 1455,
|
| 142 |
+
"sha256": "19d5b2f3b725edaeae396debb3683dda307ae239c5342e1b079a2e01328b6075"
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"path": "kernels/707b7b188f55bd9c37b6bad7708d7b38bc286b7147e6355716910e7a44f9ae82.wgsl",
|
| 146 |
+
"byteLength": 296,
|
| 147 |
+
"sha256": "707b7b188f55bd9c37b6bad7708d7b38bc286b7147e6355716910e7a44f9ae82"
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"path": "kernels/88e625463de0ca80326d53aa24f84a402f18e2e96f9ec322c6a1ac2dea47cfbf.wgsl",
|
| 151 |
+
"byteLength": 405,
|
| 152 |
+
"sha256": "88e625463de0ca80326d53aa24f84a402f18e2e96f9ec322c6a1ac2dea47cfbf"
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"path": "kernels/cd6a942d0e0231d1e8fd05a2ce362512dfaed46ceba46fc86593182de4a7a076.wgsl",
|
| 156 |
+
"byteLength": 389,
|
| 157 |
+
"sha256": "cd6a942d0e0231d1e8fd05a2ce362512dfaed46ceba46fc86593182de4a7a076"
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"path": "kernels/1be156f874ee05b766376dcddd23bbce11a1feb3c5dae1bd7d35869c8da15763.wgsl",
|
| 161 |
+
"byteLength": 389,
|
| 162 |
+
"sha256": "1be156f874ee05b766376dcddd23bbce11a1feb3c5dae1bd7d35869c8da15763"
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"path": "kernels/558739fe02ea790140335a119fbd59441cc9b64320b5bfdff4daee09d3e883e0.wgsl",
|
| 166 |
+
"byteLength": 389,
|
| 167 |
+
"sha256": "558739fe02ea790140335a119fbd59441cc9b64320b5bfdff4daee09d3e883e0"
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"path": "kernels/80b140b8a7c9a19ca722fbb3b863a3d3640c28ffb7466d84f26433a8e5cbcf17.wgsl",
|
| 171 |
+
"byteLength": 292,
|
| 172 |
+
"sha256": "80b140b8a7c9a19ca722fbb3b863a3d3640c28ffb7466d84f26433a8e5cbcf17"
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"path": "kernels/f92c5546622dc3aea6badf14abe019d854d3d9cb34a315d841bac05f9eb91066.wgsl",
|
| 176 |
+
"byteLength": 420,
|
| 177 |
+
"sha256": "f92c5546622dc3aea6badf14abe019d854d3d9cb34a315d841bac05f9eb91066"
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"path": "kernels/9070ec757115bd7c81186acd9d29ce7bc0258c77cbd02866b5dfb9c54eecd035.wgsl",
|
| 181 |
+
"byteLength": 382,
|
| 182 |
+
"sha256": "9070ec757115bd7c81186acd9d29ce7bc0258c77cbd02866b5dfb9c54eecd035"
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"path": "kernels/075bce0424ae09dc9afdefb37dbec1402758d41a459cb4473f055eca1cef73ad.wgsl",
|
| 186 |
+
"byteLength": 383,
|
| 187 |
+
"sha256": "075bce0424ae09dc9afdefb37dbec1402758d41a459cb4473f055eca1cef73ad"
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"path": "kernels/5ab3a53d6787b4ebe7eb4dae3c2da6962874d0c8cb5f23117d03473c6c76a209.wgsl",
|
| 191 |
+
"byteLength": 311,
|
| 192 |
+
"sha256": "5ab3a53d6787b4ebe7eb4dae3c2da6962874d0c8cb5f23117d03473c6c76a209"
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"path": "kernels/2bc311be08e7b4d612eda975fc9d13df28242058c479729c041a68fda6404f16.wgsl",
|
| 196 |
+
"byteLength": 562,
|
| 197 |
+
"sha256": "2bc311be08e7b4d612eda975fc9d13df28242058c479729c041a68fda6404f16"
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"path": "kernels/9d29d01524ce6f64fd00438ed7f03ce6796d0f22be1be262d6fbe43c51ffd2b6.wgsl",
|
| 201 |
+
"byteLength": 374,
|
| 202 |
+
"sha256": "9d29d01524ce6f64fd00438ed7f03ce6796d0f22be1be262d6fbe43c51ffd2b6"
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"path": "kernels/c667300f7d1060bdf36ea2a09ce4905b84ff7e9d1711fcc29c947425e0126361.wgsl",
|
| 206 |
+
"byteLength": 422,
|
| 207 |
+
"sha256": "c667300f7d1060bdf36ea2a09ce4905b84ff7e9d1711fcc29c947425e0126361"
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"path": "kernels/1d0e64f42c129b2019e2b9c0bd7ae7718b3b8bd67afa0dd6ae462619cc6401d1.wgsl",
|
| 211 |
+
"byteLength": 411,
|
| 212 |
+
"sha256": "1d0e64f42c129b2019e2b9c0bd7ae7718b3b8bd67afa0dd6ae462619cc6401d1"
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"path": "kernels/c09ba2ff8e99d3ef1fa686820ba2f6b198f2bd5f7755c8ee15dbf73344b641e8.wgsl",
|
| 216 |
+
"byteLength": 605,
|
| 217 |
+
"sha256": "c09ba2ff8e99d3ef1fa686820ba2f6b198f2bd5f7755c8ee15dbf73344b641e8"
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"path": "kernels/6b83f8154d25e77d442948180ca249a492f7d9490a41d3dda46cd9bcb2fb37fd.wgsl",
|
| 221 |
+
"byteLength": 310,
|
| 222 |
+
"sha256": "6b83f8154d25e77d442948180ca249a492f7d9490a41d3dda46cd9bcb2fb37fd"
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"path": "kernels/6619fc22f4476c11499f21aed05ce1963328fe028d204aa287d1414d3222b564.wgsl",
|
| 226 |
+
"byteLength": 1407,
|
| 227 |
+
"sha256": "6619fc22f4476c11499f21aed05ce1963328fe028d204aa287d1414d3222b564"
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"path": "kernels/55202c927e8c2e4f319a6e89a594e7524c7c36d903219eccd579c5b9b53ebf70.wgsl",
|
| 231 |
+
"byteLength": 603,
|
| 232 |
+
"sha256": "55202c927e8c2e4f319a6e89a594e7524c7c36d903219eccd579c5b9b53ebf70"
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"path": "kernels/5814d60ecf77c9336c3118f792e0bf32f3b08128ae7987f7e535598864c1a6a1.wgsl",
|
| 236 |
+
"byteLength": 409,
|
| 237 |
+
"sha256": "5814d60ecf77c9336c3118f792e0bf32f3b08128ae7987f7e535598864c1a6a1"
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"path": "kernels/83993849c704ba6fb220c0b8a0fd6a3847d253423143816289ff7e939dd204f8.wgsl",
|
| 241 |
+
"byteLength": 424,
|
| 242 |
+
"sha256": "83993849c704ba6fb220c0b8a0fd6a3847d253423143816289ff7e939dd204f8"
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"path": "kernels/cbb27a57c814432e940e67eb4f69d697bcb0099119a6bc46d85ed252de5cd9c1.wgsl",
|
| 246 |
+
"byteLength": 1455,
|
| 247 |
+
"sha256": "cbb27a57c814432e940e67eb4f69d697bcb0099119a6bc46d85ed252de5cd9c1"
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"path": "kernels/08dc45bcb86ff32fc939539b38d21b0d0d8873250e8ad9da05a42b297127cf3a.wgsl",
|
| 251 |
+
"byteLength": 1455,
|
| 252 |
+
"sha256": "08dc45bcb86ff32fc939539b38d21b0d0d8873250e8ad9da05a42b297127cf3a"
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"path": "kernels/67991f0575b151ad2e89edf7eca755a0d62686cddce367e5d3d5382518167586.wgsl",
|
| 256 |
+
"byteLength": 487,
|
| 257 |
+
"sha256": "67991f0575b151ad2e89edf7eca755a0d62686cddce367e5d3d5382518167586"
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"path": "kernels/49be6278b2b5df6dd69f6ff38e1e73dcfebc38d79f1f2694d34101dcb5c7476f.wgsl",
|
| 261 |
+
"byteLength": 1455,
|
| 262 |
+
"sha256": "49be6278b2b5df6dd69f6ff38e1e73dcfebc38d79f1f2694d34101dcb5c7476f"
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"path": "kernels/984f5c91f469bb3fdfb0d8ddfe082a64ba3dfe87fc9dc4c1db582ca1f62d5f2b.wgsl",
|
| 266 |
+
"byteLength": 1455,
|
| 267 |
+
"sha256": "984f5c91f469bb3fdfb0d8ddfe082a64ba3dfe87fc9dc4c1db582ca1f62d5f2b"
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"path": "kernels/ea188fbe79cee6ca9210d873c95f40fd3edd4c7140e6b2eb4739fedb2678e232.wgsl",
|
| 271 |
+
"byteLength": 666,
|
| 272 |
+
"sha256": "ea188fbe79cee6ca9210d873c95f40fd3edd4c7140e6b2eb4739fedb2678e232"
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"path": "kernels/0f22bffe633c43ee9cc9d88dd97725a7b0171c08b4c0e3f7fe3a53ad985f6cf4.wgsl",
|
| 276 |
+
"byteLength": 1455,
|
| 277 |
+
"sha256": "0f22bffe633c43ee9cc9d88dd97725a7b0171c08b4c0e3f7fe3a53ad985f6cf4"
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"path": "graph.json",
|
| 281 |
+
"byteLength": 3401193,
|
| 282 |
+
"sha256": "7945de65609359c334ba15f9b09aaa67f18a136f541e7d7f77d7090d7dcbeef5"
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"path": "source/config.json",
|
| 286 |
+
"byteLength": 2908,
|
| 287 |
+
"sha256": "ed1c1723241f23f7f4e23430759cbd7dcfb4103cbdfe052bfe7626b57c2615b4"
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"path": "source/LICENSE",
|
| 291 |
+
"byteLength": 11544,
|
| 292 |
+
"sha256": "bbedc3fda3305820b977265f01b8619d87570a6739de3a5582c3464840f1e57a"
|
| 293 |
+
}
|
| 294 |
+
],
|
| 295 |
+
"metadata": {
|
| 296 |
+
"repository": "Qwen/Qwen3.5-2B",
|
| 297 |
+
"revision": "15852e8c16360a2fea060d615a32b45270f8a8fc",
|
| 298 |
+
"architecture": "qwen3_5",
|
| 299 |
+
"modalities": [
|
| 300 |
+
"text",
|
| 301 |
+
"image"
|
| 302 |
+
],
|
| 303 |
+
"license": "apache-2.0",
|
| 304 |
+
"profile": "fp32-vision",
|
| 305 |
+
"imageGrid": [
|
| 306 |
+
1,
|
| 307 |
+
26,
|
| 308 |
+
38
|
| 309 |
+
],
|
| 310 |
+
"checkpointCertified": false
|
| 311 |
+
},
|
| 312 |
+
"validation": {
|
| 313 |
+
"checkpointCertified": false,
|
| 314 |
+
"hardware": []
|
| 315 |
+
}
|
| 316 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/source/LICENSE
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
Apache License
|
| 3 |
+
Version 2.0, January 2004
|
| 4 |
+
http://www.apache.org/licenses/
|
| 5 |
+
|
| 6 |
+
TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
|
| 7 |
+
|
| 8 |
+
1. Definitions.
|
| 9 |
+
|
| 10 |
+
"License" shall mean the terms and conditions for use, reproduction,
|
| 11 |
+
and distribution as defined by Sections 1 through 9 of this document.
|
| 12 |
+
|
| 13 |
+
"Licensor" shall mean the copyright owner or entity authorized by
|
| 14 |
+
the copyright owner that is granting the License.
|
| 15 |
+
|
| 16 |
+
"Legal Entity" shall mean the union of the acting entity and all
|
| 17 |
+
other entities that control, are controlled by, or are under common
|
| 18 |
+
control with that entity. For the purposes of this definition,
|
| 19 |
+
"control" means (i) the power, direct or indirect, to cause the
|
| 20 |
+
direction or management of such entity, whether by contract or
|
| 21 |
+
otherwise, or (ii) ownership of fifty percent (50%) or more of the
|
| 22 |
+
outstanding shares, or (iii) beneficial ownership of such entity.
|
| 23 |
+
|
| 24 |
+
"You" (or "Your") shall mean an individual or Legal Entity
|
| 25 |
+
exercising permissions granted by this License.
|
| 26 |
+
|
| 27 |
+
"Source" form shall mean the preferred form for making modifications,
|
| 28 |
+
including but not limited to software source code, documentation
|
| 29 |
+
source, and configuration files.
|
| 30 |
+
|
| 31 |
+
"Object" form shall mean any form resulting from mechanical
|
| 32 |
+
transformation or translation of a Source form, including but
|
| 33 |
+
not limited to compiled object code, generated documentation,
|
| 34 |
+
and conversions to other media types.
|
| 35 |
+
|
| 36 |
+
"Work" shall mean the work of authorship, whether in Source or
|
| 37 |
+
Object form, made available under the License, as indicated by a
|
| 38 |
+
copyright notice that is included in or attached to the work
|
| 39 |
+
(an example is provided in the Appendix below).
|
| 40 |
+
|
| 41 |
+
"Derivative Works" shall mean any work, whether in Source or Object
|
| 42 |
+
form, that is based on (or derived from) the Work and for which the
|
| 43 |
+
editorial revisions, annotations, elaborations, or other modifications
|
| 44 |
+
represent, as a whole, an original work of authorship. For the purposes
|
| 45 |
+
of this License, Derivative Works shall not include works that remain
|
| 46 |
+
separable from, or merely link (or bind by name) to the interfaces of,
|
| 47 |
+
the Work and Derivative Works thereof.
|
| 48 |
+
|
| 49 |
+
"Contribution" shall mean any work of authorship, including
|
| 50 |
+
the original version of the Work and any modifications or additions
|
| 51 |
+
to that Work or Derivative Works thereof, that is intentionally
|
| 52 |
+
submitted to Licensor for inclusion in the Work by the copyright owner
|
| 53 |
+
or by an individual or Legal Entity authorized to submit on behalf of
|
| 54 |
+
the copyright owner. For the purposes of this definition, "submitted"
|
| 55 |
+
means any form of electronic, verbal, or written communication sent
|
| 56 |
+
to the Licensor or its representatives, including but not limited to
|
| 57 |
+
communication on electronic mailing lists, source code control systems,
|
| 58 |
+
and issue tracking systems that are managed by, or on behalf of, the
|
| 59 |
+
Licensor for the purpose of discussing and improving the Work, but
|
| 60 |
+
excluding communication that is conspicuously marked or otherwise
|
| 61 |
+
designated in writing by the copyright owner as "Not a Contribution."
|
| 62 |
+
|
| 63 |
+
"Contributor" shall mean Licensor and any individual or Legal Entity
|
| 64 |
+
on behalf of whom a Contribution has been received by Licensor and
|
| 65 |
+
subsequently incorporated within the Work.
|
| 66 |
+
|
| 67 |
+
2. Grant of Copyright License. Subject to the terms and conditions of
|
| 68 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 69 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 70 |
+
copyright license to reproduce, prepare Derivative Works of,
|
| 71 |
+
publicly display, publicly perform, sublicense, and distribute the
|
| 72 |
+
Work and such Derivative Works in Source or Object form.
|
| 73 |
+
|
| 74 |
+
3. Grant of Patent License. Subject to the terms and conditions of
|
| 75 |
+
this License, each Contributor hereby grants to You a perpetual,
|
| 76 |
+
worldwide, non-exclusive, no-charge, royalty-free, irrevocable
|
| 77 |
+
(except as stated in this section) patent license to make, have made,
|
| 78 |
+
use, offer to sell, sell, import, and otherwise transfer the Work,
|
| 79 |
+
where such license applies only to those patent claims licensable
|
| 80 |
+
by such Contributor that are necessarily infringed by their
|
| 81 |
+
Contribution(s) alone or by combination of their Contribution(s)
|
| 82 |
+
with the Work to which such Contribution(s) was submitted. If You
|
| 83 |
+
institute patent litigation against any entity (including a
|
| 84 |
+
cross-claim or counterclaim in a lawsuit) alleging that the Work
|
| 85 |
+
or a Contribution incorporated within the Work constitutes direct
|
| 86 |
+
or contributory patent infringement, then any patent licenses
|
| 87 |
+
granted to You under this License for that Work shall terminate
|
| 88 |
+
as of the date such litigation is filed.
|
| 89 |
+
|
| 90 |
+
4. Redistribution. You may reproduce and distribute copies of the
|
| 91 |
+
Work or Derivative Works thereof in any medium, with or without
|
| 92 |
+
modifications, and in Source or Object form, provided that You
|
| 93 |
+
meet the following conditions:
|
| 94 |
+
|
| 95 |
+
(a) You must give any other recipients of the Work or
|
| 96 |
+
Derivative Works a copy of this License; and
|
| 97 |
+
|
| 98 |
+
(b) You must cause any modified files to carry prominent notices
|
| 99 |
+
stating that You changed the files; and
|
| 100 |
+
|
| 101 |
+
(c) You must retain, in the Source form of any Derivative Works
|
| 102 |
+
that You distribute, all copyright, patent, trademark, and
|
| 103 |
+
attribution notices from the Source form of the Work,
|
| 104 |
+
excluding those notices that do not pertain to any part of
|
| 105 |
+
the Derivative Works; and
|
| 106 |
+
|
| 107 |
+
(d) If the Work includes a "NOTICE" text file as part of its
|
| 108 |
+
distribution, then any Derivative Works that You distribute must
|
| 109 |
+
include a readable copy of the attribution notices contained
|
| 110 |
+
within such NOTICE file, excluding those notices that do not
|
| 111 |
+
pertain to any part of the Derivative Works, in at least one
|
| 112 |
+
of the following places: within a NOTICE text file distributed
|
| 113 |
+
as part of the Derivative Works; within the Source form or
|
| 114 |
+
documentation, if provided along with the Derivative Works; or,
|
| 115 |
+
within a display generated by the Derivative Works, if and
|
| 116 |
+
wherever such third-party notices normally appear. The contents
|
| 117 |
+
of the NOTICE file are for informational purposes only and
|
| 118 |
+
do not modify the License. You may add Your own attribution
|
| 119 |
+
notices within Derivative Works that You distribute, alongside
|
| 120 |
+
or as an addendum to the NOTICE text from the Work, provided
|
| 121 |
+
that such additional attribution notices cannot be construed
|
| 122 |
+
as modifying the License.
|
| 123 |
+
|
| 124 |
+
You may add Your own copyright statement to Your modifications and
|
| 125 |
+
may provide additional or different license terms and conditions
|
| 126 |
+
for use, reproduction, or distribution of Your modifications, or
|
| 127 |
+
for any such Derivative Works as a whole, provided Your use,
|
| 128 |
+
reproduction, and distribution of the Work otherwise complies with
|
| 129 |
+
the conditions stated in this License.
|
| 130 |
+
|
| 131 |
+
5. Submission of Contributions. Unless You explicitly state otherwise,
|
| 132 |
+
any Contribution intentionally submitted for inclusion in the Work
|
| 133 |
+
by You to the Licensor shall be under the terms and conditions of
|
| 134 |
+
this License, without any additional terms or conditions.
|
| 135 |
+
Notwithstanding the above, nothing herein shall supersede or modify
|
| 136 |
+
the terms of any separate license agreement you may have executed
|
| 137 |
+
with Licensor regarding such Contributions.
|
| 138 |
+
|
| 139 |
+
6. Trademarks. This License does not grant permission to use the trade
|
| 140 |
+
names, trademarks, service marks, or product names of the Licensor,
|
| 141 |
+
except as required for reasonable and customary use in describing the
|
| 142 |
+
origin of the Work and reproducing the content of the NOTICE file.
|
| 143 |
+
|
| 144 |
+
7. Disclaimer of Warranty. Unless required by applicable law or
|
| 145 |
+
agreed to in writing, Licensor provides the Work (and each
|
| 146 |
+
Contributor provides its Contributions) on an "AS IS" BASIS,
|
| 147 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
|
| 148 |
+
implied, including, without limitation, any warranties or conditions
|
| 149 |
+
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
|
| 150 |
+
PARTICULAR PURPOSE. You are solely responsible for determining the
|
| 151 |
+
appropriateness of using or redistributing the Work and assume any
|
| 152 |
+
risks associated with Your exercise of permissions under this License.
|
| 153 |
+
|
| 154 |
+
8. Limitation of Liability. In no event and under no legal theory,
|
| 155 |
+
whether in tort (including negligence), contract, or otherwise,
|
| 156 |
+
unless required by applicable law (such as deliberate and grossly
|
| 157 |
+
negligent acts) or agreed to in writing, shall any Contributor be
|
| 158 |
+
liable to You for damages, including any direct, indirect, special,
|
| 159 |
+
incidental, or consequential damages of any character arising as a
|
| 160 |
+
result of this License or out of the use or inability to use the
|
| 161 |
+
Work (including but not limited to damages for loss of goodwill,
|
| 162 |
+
work stoppage, computer failure or malfunction, or any and all
|
| 163 |
+
other commercial damages or losses), even if such Contributor
|
| 164 |
+
has been advised of the possibility of such damages.
|
| 165 |
+
|
| 166 |
+
9. Accepting Warranty or Additional Liability. While redistributing
|
| 167 |
+
the Work or Derivative Works thereof, You may choose to offer,
|
| 168 |
+
and charge a fee for, acceptance of support, warranty, indemnity,
|
| 169 |
+
or other liability obligations and/or rights consistent with this
|
| 170 |
+
License. However, in accepting such obligations, You may act only
|
| 171 |
+
on Your own behalf and on Your sole responsibility, not on behalf
|
| 172 |
+
of any other Contributor, and only if You agree to indemnify,
|
| 173 |
+
defend, and hold each Contributor harmless for any liability
|
| 174 |
+
incurred by, or claims asserted against, such Contributor by reason
|
| 175 |
+
of your accepting any such warranty or additional liability.
|
| 176 |
+
|
| 177 |
+
END OF TERMS AND CONDITIONS
|
| 178 |
+
|
| 179 |
+
APPENDIX: How to apply the Apache License to your work.
|
| 180 |
+
|
| 181 |
+
To apply the Apache License to your work, attach the following
|
| 182 |
+
boilerplate notice, with the fields enclosed by brackets "[]"
|
| 183 |
+
replaced with your own identifying information. (Don't include
|
| 184 |
+
the brackets!) The text should be enclosed in the appropriate
|
| 185 |
+
comment syntax for the file format. We also recommend that a
|
| 186 |
+
file or class name and description of purpose be included on the
|
| 187 |
+
same "printed page" as the copyright notice for easier
|
| 188 |
+
identification within third-party archives.
|
| 189 |
+
|
| 190 |
+
Copyright 2026 Alibaba Cloud
|
| 191 |
+
|
| 192 |
+
Licensed under the Apache License, Version 2.0 (the "License");
|
| 193 |
+
you may not use this file except in compliance with the License.
|
| 194 |
+
You may obtain a copy of the License at
|
| 195 |
+
|
| 196 |
+
http://www.apache.org/licenses/LICENSE-2.0
|
| 197 |
+
|
| 198 |
+
Unless required by applicable law or agreed to in writing, software
|
| 199 |
+
distributed under the License is distributed on an "AS IS" BASIS,
|
| 200 |
+
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
| 201 |
+
See the License for the specific language governing permissions and
|
| 202 |
+
limitations under the License.
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/source/config.json
ADDED
|
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3_5ForConditionalGeneration"
|
| 4 |
+
],
|
| 5 |
+
"image_token_id": 248056,
|
| 6 |
+
"model_type": "qwen3_5",
|
| 7 |
+
"text_config": {
|
| 8 |
+
"attention_bias": false,
|
| 9 |
+
"attention_dropout": 0.0,
|
| 10 |
+
"attn_output_gate": true,
|
| 11 |
+
"dtype": "bfloat16",
|
| 12 |
+
"eos_token_id": 248044,
|
| 13 |
+
"full_attention_interval": 4,
|
| 14 |
+
"head_dim": 256,
|
| 15 |
+
"hidden_act": "silu",
|
| 16 |
+
"hidden_size": 2048,
|
| 17 |
+
"initializer_range": 0.02,
|
| 18 |
+
"intermediate_size": 6144,
|
| 19 |
+
"layer_types": [
|
| 20 |
+
"linear_attention",
|
| 21 |
+
"linear_attention",
|
| 22 |
+
"linear_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"linear_attention",
|
| 25 |
+
"linear_attention",
|
| 26 |
+
"linear_attention",
|
| 27 |
+
"full_attention",
|
| 28 |
+
"linear_attention",
|
| 29 |
+
"linear_attention",
|
| 30 |
+
"linear_attention",
|
| 31 |
+
"full_attention",
|
| 32 |
+
"linear_attention",
|
| 33 |
+
"linear_attention",
|
| 34 |
+
"linear_attention",
|
| 35 |
+
"full_attention",
|
| 36 |
+
"linear_attention",
|
| 37 |
+
"linear_attention",
|
| 38 |
+
"linear_attention",
|
| 39 |
+
"full_attention",
|
| 40 |
+
"linear_attention",
|
| 41 |
+
"linear_attention",
|
| 42 |
+
"linear_attention",
|
| 43 |
+
"full_attention"
|
| 44 |
+
],
|
| 45 |
+
"linear_conv_kernel_dim": 4,
|
| 46 |
+
"linear_key_head_dim": 128,
|
| 47 |
+
"linear_num_key_heads": 16,
|
| 48 |
+
"linear_num_value_heads": 16,
|
| 49 |
+
"linear_value_head_dim": 128,
|
| 50 |
+
"max_position_embeddings": 262144,
|
| 51 |
+
"mlp_only_layers": [],
|
| 52 |
+
"model_type": "qwen3_5_text",
|
| 53 |
+
"mtp_num_hidden_layers": 1,
|
| 54 |
+
"mtp_use_dedicated_embeddings": false,
|
| 55 |
+
"num_attention_heads": 8,
|
| 56 |
+
"num_hidden_layers": 24,
|
| 57 |
+
"num_key_value_heads": 2,
|
| 58 |
+
"rms_norm_eps": 1e-06,
|
| 59 |
+
"tie_word_embeddings": true,
|
| 60 |
+
"use_cache": true,
|
| 61 |
+
"vocab_size": 248320,
|
| 62 |
+
"mamba_ssm_dtype": "float32",
|
| 63 |
+
"rope_parameters": {
|
| 64 |
+
"mrope_interleaved": true,
|
| 65 |
+
"mrope_section": [
|
| 66 |
+
11,
|
| 67 |
+
11,
|
| 68 |
+
10
|
| 69 |
+
],
|
| 70 |
+
"rope_type": "default",
|
| 71 |
+
"rope_theta": 10000000,
|
| 72 |
+
"partial_rotary_factor": 0.25
|
| 73 |
+
}
|
| 74 |
+
},
|
| 75 |
+
"tie_word_embeddings": true,
|
| 76 |
+
"transformers_version": "4.57.0.dev0",
|
| 77 |
+
"video_token_id": 248057,
|
| 78 |
+
"vision_config": {
|
| 79 |
+
"deepstack_visual_indexes": [],
|
| 80 |
+
"depth": 24,
|
| 81 |
+
"hidden_act": "gelu_pytorch_tanh",
|
| 82 |
+
"hidden_size": 1024,
|
| 83 |
+
"in_channels": 3,
|
| 84 |
+
"initializer_range": 0.02,
|
| 85 |
+
"intermediate_size": 4096,
|
| 86 |
+
"model_type": "qwen3_5",
|
| 87 |
+
"num_heads": 16,
|
| 88 |
+
"num_position_embeddings": 2304,
|
| 89 |
+
"out_hidden_size": 2048,
|
| 90 |
+
"patch_size": 16,
|
| 91 |
+
"spatial_merge_size": 2,
|
| 92 |
+
"temporal_patch_size": 2
|
| 93 |
+
},
|
| 94 |
+
"vision_end_token_id": 248054,
|
| 95 |
+
"vision_start_token_id": 248053
|
| 96 |
+
}
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00000.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5ff01e44eac7d7a30d3779fe7970f1ec882ae3a807ded59b8c25b203b57bd746
|
| 3 |
+
size 67108864
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00001.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f698d7140ff322989ac4a6aa5c05976740689c7129be1f179362c94bef2730f
|
| 3 |
+
size 67108864
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00002.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a05e5a106c236dedfca3a681806500e8d9f518767ec432ffa7a53a0061e94f8
|
| 3 |
+
size 67108864
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00003.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da533794b3c088224ed91b19ce038c0b07acbfa7b01eeb5b6b1e0d8bc39e5bc8
|
| 3 |
+
size 67108864
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00004.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b95eb7e977231963dd5335fab05dc5d111c3c5c716bc0bac3e0f48af84b5cf18
|
| 3 |
+
size 67108864
|
qwen35-2b-multimodal-fp32-three-grid-state-alias-token-major-v2-v5/vision/416x608/weights/00005.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17c20eaa5be55412038d213e9820d01a4f1f822678a1fbe5d7df20b5012b546e
|
| 3 |
+
size 67108864
|