Skip to content

Comments

Create PagedExtendAttention class to use Wave KVCache layout#2586

Draft
aviator19941 wants to merge 29 commits intomainfrom
wave_kv_cache_layout
Draft

Create PagedExtendAttention class to use Wave KVCache layout#2586
aviator19941 wants to merge 29 commits intomainfrom
wave_kv_cache_layout

Conversation

@aviator19941
Copy link
Collaborator

Refactor of #2534

Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
…lConfig

Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
… class is same logic

Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
@github-actions
Copy link
Contributor

github-actions bot commented Nov 1, 2025

Coverage report

Click to see where and how coverage changed

FileStatementsMissingCoverageCoverage
(new stmts)
Lines missing
  sharktank/sharktank/examples
  export_paged_llm_v1.py 126-127
  sharktank/sharktank/kernels/wave
  utils.py 103-115, 318, 330-344, 350-375, 386-404
  sharktank/sharktank/kernels/wave/templates
  extend_attention_kernel.py 155-161, 216-222
  sharktank/sharktank/layers
  __init__.py
  ffn_block.py
  linear.py
  norm.py
  paged_attention.py 164-200, 419-437, 448-477, 493-521, 532-562, 729, 1223-1231, 1234, 1244, 1261, 1278, 1305-1321, 1354-1362, 1399-1415, 1453-1491, 1524-1545
  paged_llama_attention_block.py 94-101, 165
  token_embedding.py
  sharktank/sharktank/layers/configs
  llm_configs.py
  sharktank/sharktank/models/llm
  export.py
  llm.py
  sharktank/sharktank/ops
  attention_impls.py 257-290
  signatures.py 904
  sharktank/sharktank/types
  tensors.py
  sharktank/sharktank/utils
  create_cache.py 21, 28-29, 47, 54
  testing.py 903, 907-911, 942-955, 965, 979-999, 1004-1007, 1014-1038, 1055-1070, 1074-1079
  sharktank/tests/kernels/wave
  extend_attention_test.py 208, 249-315, 333-340, 355-375, 388-395, 407-414, 424-429, 454-523, 542-620
Project Total  

This report was generated by python-coverage-comment-action

Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Signed-off-by: aviator19941 <avinash.sharma@amd.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant