Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion it/test_all_tracks_and_challenges.py
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@


class TestTrackRepository:
skip_tracks = ["elastic/logs", "elastic/security", "k8s_metrics", "sql", "elser-ingest-speedtest", "msmarco-v2-vector"]
skip_tracks = ["elastic/logs", "elastic/security", "k8s_metrics", "sql", "elser-ingest-speedtest", "msmarco-v2-vector", "openai_vector"]
disable_assertions = {
"http_logs": ["append-no-conflicts", "runtime-fields"],
"nyc_taxis": ["update-aggs-only"],
Expand Down
1 change: 1 addition & 0 deletions openai_vector/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -51,6 +51,7 @@ This track accepts the following parameters with Rally 0.8.0+ using `--track-par
- standalone_search_clients (default: 8)
- standalone_search_iterations (default: 10000)
- vector_index_type (default: "hnsw"): The index kind for storing the vectors.
- vector_ops (default: [[10, 20, 0], [10, 20, 1], [10, 20, 2], [10, 50, 1], [10, 50, 2], [10, 100, 1], [100, 120, 1], [100, 120, 2], [100, 200, 1], [100, 200, 2], [100, 500, 1], [100, 500, 2]]): The vector search operations, formattied [k, num_candidates, oversample], where `oversample` indicates the ratio of extra `k` to gather and then rescore.

### License

Expand Down
75 changes: 35 additions & 40 deletions openai_vector/challenges/default.json
Original file line number Diff line number Diff line change
Expand Up @@ -42,61 +42,56 @@
"retry-until-success": true,
"include-in-reporting": false
}
},
}
{# serverless-post-ingest-sleep-marker-start #}{%- if post_ingest_sleep|default(false) -%}
{
,{
"name": "post-ingest-sleep",
"operation": {
"operation-type": "sleep",
"duration": {{ post_ingest_sleep_duration|default(30) }}
}
},
}
{%- endif -%}{# serverless-post-ingest-sleep-marker-end #}
{%- for i in range(p_search_ops|length) %},
{
"name": "standalone-search-knn-10-100-single-client",
"operation": "knn-search-10-100",
"warmup-iterations": 100,
"iterations": {{ standalone_search_iterations | default(10000) | int }}
},
{
"name": "standalone-knn-search-100-1000-single-client",
"operation": "knn-search-100-1000",
"warmup-iterations": 100,
{%- if p_search_ops[i][2] > 0 -%}
"name": "standalone-search-knn-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}-single-client",
"operation": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}"
{%- else -%}
"name": "standalone-search-knn-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-single-client",
"operation": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}"
{%- endif -%},
"warmup-iterations": 1000,
"iterations": {{ standalone_search_iterations | default(10000) | int }}
},
{
"name": "standalone-search-knn-10-100-multiple-clients",
"operation": "knn-search-10-100",
"warmup-iterations": 100,
{%- if p_search_ops[i][2] > 0 -%}
"name": "standalone-search-knn-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}-multiple-clients",
"operation": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}"
{%- else -%}
"name": "standalone-search-knn-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-multiple-clients",
"operation": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}"
{%- endif -%},
"warmup-iterations": 1000,
"clients": {{ standalone_search_clients | default(8) | int }},
"iterations": {{ standalone_search_iterations | default(10000) | int }}
},
}
{%- endfor %},
{
"name": "standalone-search-knn-100-1000-multiple-clients",
"operation": "knn-search-100-1000",
"warmup-iterations": 100,
"clients": {{ standalone_search_clients | default(8) | int }},
"iterations": {{ standalone_search_iterations | default(10000) | int }}
},
"name": "parallel-documents-indexing-bulk",
"operation": "parallel-documents-indexing",
"warmup-time-period": 60,
"clients": {{ parallel_indexing_bulk_clients | default(1) | int }},
"target-throughput": {{ parallel_indexing_bulk_target_throughput | default(1) | int }}
}
{%- for i in range(p_search_ops|length) %},
{
"parallel": {
"tasks": [
{
"name": "parallel-documents-indexing-bulk",
"operation": "parallel-documents-indexing",
"clients": {{ parallel_indexing_bulk_clients | default(1) | int }},
"time-period": {{ parallel_indexing_time_period | default(1800) | int }},
"target-throughput": {{ parallel_indexing_bulk_target_throughput | default(1) | int }}
},
{
"name": "parallel-documents-indexing-search-knn-10-100",
"operation": "knn-search-10-100",
"clients": {{ parallel_indexing_search_clients | default(3) | int }},
"time-period": {{ parallel_indexing_time_period | default(1800) | int }},
"target-throughput": {{ parallel_indexing_search_target_throughput | default(100) | int }}
}
]
}
{%- if p_search_ops[i][2] > 0 -%}
"operation": "knn-recall-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}"
{%- else -%}
"operation": "knn-recall-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}"
{%- endif -%}
}
{%- endfor %}
]
}
4 changes: 2 additions & 2 deletions openai_vector/index-vectors-only-mapping.json
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
"settings": {
{# non-serverless-index-settings-marker-start #}{%- if build_flavor != "serverless" or serverless_operator == true -%}
{% if preload_pagecache %}
"index.store.preload": [ "vec", "vex", "vem"],
"index.store.preload": [ "vec", "vex", "vem", "veq", "veqm", "veb", "vebm"],
{% endif %}
"index.number_of_shards": {{number_of_shards | default(1)}},
"index.number_of_replicas": {{number_of_replicas | default(0)}}
Expand All @@ -21,7 +21,7 @@
"index": true,
"similarity": "dot_product",
"index_options": {
"type": {{ vector_index_type | default("hnsw") | tojson }}
"type": {{ vector_index_type | default("int8_hnsw") | tojson }}
}
}
}
Expand Down
29 changes: 29 additions & 0 deletions openai_vector/index-vectors-only-with-docid-mapping.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
{
"settings": {
{# non-serverless-index-settings-marker-start #}{%- if build_flavor != "serverless" or serverless_operator == true -%}
{% if preload_pagecache %}
"index.store.preload": [ "vec", "vex", "vem", "veq", "veqm", "veb", "vebm"],
{% endif %}
"index.number_of_shards": {{number_of_shards | default(1)}},
"index.number_of_replicas": {{number_of_replicas | default(0)}}
{%- endif -%}{# non-serverless-index-settings-marker-end #}
},
"mappings": {
"dynamic": false,
"properties": {
"docid": {
"type": "keyword"
},
"emb": {
"type": "dense_vector",
"element_type": "float",
"dims": 1536,
"index": true,
"similarity": "dot_product",
"index_options": {
"type": {{ vector_index_type | default("int8_hnsw") | tojson }}
}
}
}
}
}
4 changes: 2 additions & 2 deletions openai_vector/index-vectors-with-text-mapping.json
Original file line number Diff line number Diff line change
Expand Up @@ -2,7 +2,7 @@
"settings": {
{# non-serverless-index-settings-marker-start #}{%- if build_flavor != "serverless" or serverless_operator == true -%}
{% if preload_pagecache %}
"index.store.preload": [ "vec", "vex", "vem"],
"index.store.preload": [ "vec", "vex", "vem", "veq", "veqm", "veb", "vebm"],
{% endif %}
"index.number_of_shards": {{number_of_shards | default(1)}},
"index.number_of_replicas": {{number_of_replicas | default(0)}}
Expand All @@ -26,7 +26,7 @@
"index": true,
"similarity": "dot_product",
"index_options": {
"type": {{ vector_index_type | default("hnsw") | tojson }}
"type": {{ vector_index_type | default("int8_hnsw") | tojson }}
}
}
}
Expand Down
Binary file added openai_vector/open_ai_true_top_1000.json.bz2
Binary file not shown.
31 changes: 22 additions & 9 deletions openai_vector/operations/default.json
Original file line number Diff line number Diff line change
Expand Up @@ -24,18 +24,31 @@
"corpora": "openai-parallel-indexing",
"bulk-size": {{parallel_indexing_bulk_size | default(500)}},
"ingest-percentage": {{parallel_indexing_ingest_percentage | default(100)}}
},
}
{%- set p_search_ops = (search_ops | default([(10, 20, 0), (10, 20, 1), (10, 20, 2), (10, 50, 1), (10, 50, 2), (10, 100, 1), (100, 120, 1), (100, 120, 2), (100, 200, 1), (100, 200, 2), (100, 500, 1), (100, 500, 2)]))%}
{%- for i in range(p_search_ops|length) %},
{
"name": "knn-search-10-100",
{%- if p_search_ops[i][2] > 0 -%}
"name": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}"
{%- else -%}
"name": "knn-search-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}"
{%- endif -%},
"operation-type": "search",
"param-source": "knn-param-source",
"k": 10,
"num-candidates": 100
"k": {{p_search_ops[i][0]}},
"num-candidates": {{p_search_ops[i][1]}},
"oversample": {{p_search_ops[i][2]}}
},
{
"name": "knn-search-100-1000",
"operation-type": "search",
"param-source": "knn-param-source",
"k": 100,
"num-candidates": 1000
{%- if p_search_ops[i][2] > 0 -%}
"name": "knn-recall-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}-{{p_search_ops[i][2]}}"
{%- else -%}
"name": "knn-recall-{{p_search_ops[i][0]}}-{{p_search_ops[i][1]}}"
{%- endif -%},
"operation-type": "knn-recall",
"param-source": "knn-recall-param-source",
"k": {{p_search_ops[i][0]}},
"num-candidates": {{p_search_ops[i][1]}},
"oversample": {{p_search_ops[i][2]}}
}
{%- endfor %}
2 changes: 1 addition & 1 deletion openai_vector/track.json
Original file line number Diff line number Diff line change
Expand Up @@ -6,7 +6,7 @@
"indices": [
{
"name": "openai",
"body": "index-{{ mapping_type | default("vectors-only") }}-mapping.json"
"body": "index-{{ mapping_type | default("vectors-only-with-docid") }}-mapping.json"
}
],
"corpora": [
Expand Down
Loading
Loading