[Model] Support Qwen3.8-Flash-Next (#53896)

Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
Signed-off-by: andyluo7 <andy.luo@amd.com>
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Signed-off-by: Harry Huang <huanghaoyan.hhy@alibaba-inc.com>
Signed-off-by: Thien Tran <gau.nernst@yahoo.com.sg>
Signed-off-by: Isotr0py <Isotr0py@outlook.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai>
Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
Co-authored-by: zjy0516 <riverclouds.zhu@qq.com>
Co-authored-by: KungYork <firelikest@163.com>
Co-authored-by: Thien Tran <gau.nernst@yahoo.com.sg>
Co-authored-by: Yifan Qiao <yifanqiao@inferact.ai>
Co-authored-by: andyluo7 <43718156+andyluo7@users.noreply.github.com>
Co-authored-by: Isotr0py <mozf@inferact.ai>
Co-authored-by: andyluo7 <andy.luo@amd.com>
Co-authored-by: OpenAI Codex <noreply@openai.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
Co-authored-by: Vadim Gimpelson <156319763+vadiklyutiy@users.noreply.github.com>
Co-authored-by: Isotr0py <Isotr0py@outlook.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
Harry Huang
2026-08-31 05:57:56 +00:00
committed by GitHub
co-authored by zjy0516 KungYork Thien Tran Yifan Qiao andyluo7 Isotr0py andyluo7 OpenAI Codex lishunyang12 Vadim Gimpelson Isotr0py Lucas Wilkinson Claude Opus 5 OpenAI Codex
parent 687db59744
commit e126687a9a
124 changed files with 19424 additions and 450 deletions
+36
View File
@@ -132,6 +132,42 @@ def test_block_tables_apply_staged_writes_single_group():
)
def test_block_tables_skip_custom_slot_mapping_groups():
device = torch.device("cuda")
block_tables = BlockTables(
block_sizes=[8, 262144],
max_num_reqs=1,
max_num_batched_tokens=4,
max_num_blocks_per_group=[1, 1],
device=device,
kernel_block_sizes=[8, 262144],
slot_mapping_enabled=[False, True],
)
block_tables.append_block_ids(
req_index=0,
new_block_ids=([7], [12]),
overwrite=True,
)
block_tables.apply_staged_writes()
idx_mapping = torch.tensor([0], dtype=torch.int32, device=device)
query_start_loc = torch.tensor([0, 2], dtype=torch.int32, device=device)
positions = torch.tensor([153797, 165757], dtype=torch.int64, device=device)
slot_mappings = block_tables.compute_slot_mappings(
idx_mapping,
query_start_loc,
positions,
num_tokens_padded=2,
)
torch.accelerator.synchronize()
assert slot_mappings[0].tolist() == [-1, -1]
assert slot_mappings[1].tolist() == [
12 * 262144 + 153797,
12 * 262144 + 165757,
]
@pytest.mark.parametrize("cp_rank", range(4))
def test_dcp_slot_mapping_with_smaller_kernel_blocks(cp_rank: int):
"""DCP interleave is expressed in logical-block token coordinates."""