All workflows

Showing runs from all workflows

2500+ workflow runs

Implements the kv_save_blocks_total counter metric (tpu_raiden_kv_save_blocks_total{dp_rank, tier, outcome}) to track block save outcomes across KV cache offload tiers:
completed successfully:
GitHub Actions Scan#6361:Pull request#1371synchronizebycopybara-service[bot]Bot
Oct 29, 202612s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Presubmit Benchmarks#2672:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 202610m 14s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6338:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 202647s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Unit Tests#2153:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 20267m 39s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
GitHub Actions Scan#6360:Pull request#1395synchronizebycopybara-service[bot]Bot
Oct 29, 20269s
Index active transfer schedules by block ID in RegisterActivePlan so KVCacheManagerBase::GetBlockChunks scales linearly O(N) instead of quadratically O(N²).
completed successfully:
Presubmit Benchmarks#2671:Pull request#1397openedbycopybara-service[bot]Bot
test_1008056720
Oct 29, 202610m 17s
Index active transfer schedules by block ID in RegisterActivePlan so KVCacheManagerBase::GetBlockChunks scales linearly O(N) instead of quadratically O(N²).
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6337:Pull request#1397openedbycopybara-service[bot]Bot
test_1008056720
Oct 29, 202645s
Index active transfer schedules by block ID in RegisterActivePlan so KVCacheManagerBase::GetBlockChunks scales linearly O(N) instead of quadratically O(N²).
completed successfully:
GitHub Actions Scan#6359:Pull request#1397openedbycopybara-service[bot]Bot
Oct 29, 202616s
Index active transfer schedules by block ID in RegisterActivePlan so KVCacheManagerBase::GetBlockChunks scales linearly O(N) instead of quadratically O(N²).
completed successfully:
Unit Tests#2152:Pull request#1397openedbycopybara-service[bot]Bot
test_1008056720
Oct 29, 20265m 22s
Add the kv_evicted_blocks_total (tpu_raiden_kv_evicted_blocks_total) counter metric to tpu_raiden telemetry and instrument KVCacheStore to record host DRAM LRU evictions labeled by {dp_rank}.
completed successfully:
Presubmit Benchmarks#2670:Pull request#1396openedbycopybara-service[bot]Bot
test_999559864
Oct 29, 202610m 37s
Add the kv_evicted_blocks_total (tpu_raiden_kv_evicted_blocks_total) counter metric to tpu_raiden telemetry and instrument KVCacheStore to record host DRAM LRU evictions labeled by {dp_rank}.
completed successfully:
GitHub Actions Scan#6358:Pull request#1396openedbycopybara-service[bot]Bot
Oct 29, 202612s
Add the kv_evicted_blocks_total (tpu_raiden_kv_evicted_blocks_total) counter metric to tpu_raiden telemetry and instrument KVCacheStore to record host DRAM LRU evictions labeled by {dp_rank}.
completed successfully:
Unit Tests#2151:Pull request#1396openedbycopybara-service[bot]Bot
test_999559864
Oct 29, 20266m 56s
Add the kv_evicted_blocks_total (tpu_raiden_kv_evicted_blocks_total) counter metric to tpu_raiden telemetry and instrument KVCacheStore to record host DRAM LRU evictions labeled by {dp_rank}.
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6336:Pull request#1396openedbycopybara-service[bot]Bot
test_999559864
Oct 29, 202637s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Presubmit Benchmarks#2669:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 20269m 53s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Unit Tests#2150:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 20264m 24s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6335:Pull request#1395synchronizebycopybara-service[bot]Bot
test_1009082311
Oct 29, 202636s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
GitHub Actions Scan#6357:Pull request#1395synchronizebycopybara-service[bot]Bot
Oct 29, 20269s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
GitHub Actions Scan#6356:Pull request#1395openedbycopybara-service[bot]Bot
Oct 29, 20268s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Presubmit Benchmarks#2668:Pull request#1395openedbycopybara-service[bot]Bot
test_1009082311
Oct 29, 202610m 8s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Unit Tests#2149:Pull request#1395openedbycopybara-service[bot]Bot
test_1009082311
Oct 29, 20266m 50s
Report a 4-part breakdown of host DRAM usage in bytes through the tpu_raiden_host_memory_breakdown_bytes gauge ({component}), registered in cl/1011877945:
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6334:Pull request#1395openedbycopybara-service[bot]Bot
test_1009082311
Oct 29, 202642s
Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN
completed successfully:
Presubmit Benchmarks#2667:Pull request#1394openedbycopybara-service[bot]Bot
test_1009546175
Oct 29, 20269m 6s
Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN
completed successfully:
GitHub Actions Scan#6355:Pull request#1394openedbycopybara-service[bot]Bot
Oct 29, 202613s
Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN
completed successfully:
Google GitHub Admin: Actions Workflow Security Scan#6333:Pull request#1394openedbycopybara-service[bot]Bot
test_1009546175
Oct 29, 202648s
Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN
completed successfully:
Unit Tests#2148:Pull request#1394openedbycopybara-service[bot]Bot
test_1009546175
Oct 29, 20265m 27s

Navigation Menu

All workflows

2500+ workflow runs

Implements the `kv_save_blocks_total` counter metric (`tpu_raiden_kv_save_blocks_total{dp_rank, tier, outcome}`) to track block save outcomes across KV cache offload tiers:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Index active transfer schedules by block ID in `RegisterActivePlan` so `KVCacheManagerBase::GetBlockChunks` scales linearly `O(N)` instead of quadratically `O(N²)`.

Index active transfer schedules by block ID in `RegisterActivePlan` so `KVCacheManagerBase::GetBlockChunks` scales linearly `O(N)` instead of quadratically `O(N²)`.

Index active transfer schedules by block ID in `RegisterActivePlan` so `KVCacheManagerBase::GetBlockChunks` scales linearly `O(N)` instead of quadratically `O(N²)`.

Index active transfer schedules by block ID in `RegisterActivePlan` so `KVCacheManagerBase::GetBlockChunks` scales linearly `O(N)` instead of quadratically `O(N²)`.

Add the `kv_evicted_blocks_total` (`tpu_raiden_kv_evicted_blocks_total`) counter metric to `tpu_raiden` telemetry and instrument `KVCacheStore` to record host DRAM LRU evictions labeled by `{dp_rank}`.

Add the `kv_evicted_blocks_total` (`tpu_raiden_kv_evicted_blocks_total`) counter metric to `tpu_raiden` telemetry and instrument `KVCacheStore` to record host DRAM LRU evictions labeled by `{dp_rank}`.

Add the `kv_evicted_blocks_total` (`tpu_raiden_kv_evicted_blocks_total`) counter metric to `tpu_raiden` telemetry and instrument `KVCacheStore` to record host DRAM LRU evictions labeled by `{dp_rank}`.

Add the `kv_evicted_blocks_total` (`tpu_raiden_kv_evicted_blocks_total`) counter metric to `tpu_raiden` telemetry and instrument `KVCacheStore` to record host DRAM LRU evictions labeled by `{dp_rank}`.

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Report a 4-part breakdown of host DRAM usage in bytes through the `tpu_raiden_host_memory_breakdown_bytes` gauge (`{component}`), registered in cl/1011877945:

Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN

Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN

Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN

Replace deprecated TF_ASSERT_OK_AND_ASSIGN with ASSERT_OK_AND_ASSIGN

ProTip! You can narrow down the results and go further in time using the filter or the other filters available.

Workflow runs

2500+ workflow runs

ProTip! You can narrow down the results and go further in time using the created: filter or the other filters available.

ProTip! You can narrow down the results and go further in time using the filter or the other filters available.