Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

enhance: [2.4] support recall estimation #38064

Merged
merged 2 commits into from
Dec 11, 2024
Merged

enhance: [2.4] support recall estimation #38064

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a921b00
enhance: support recall estimation
chasingegg Nov 26, 2024
8e6853c
update
chasingegg Dec 11, 2024
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: 2 additions & 0 deletions internal/proto/internal.proto
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,6 +122,7 @@ message SearchRequest {
int64 offset = 21;
common.ConsistencyLevel consistency_level = 22;
bool is_topk_reduce = 26;
bool is_recall_evaluation = 27;
}

message SubSearchResults {
Expand Down Expand Up @@ -158,6 +159,7 @@ message SearchResults {
bool is_advanced = 16;
int64 all_search_count = 17;
bool is_topk_reduce = 18;
bool is_recall_evaluation = 19;
}

message CostAggregation {
Expand Down
41 changes: 30 additions & 11 deletions internal/proxy/impl.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2967,12 +2967,13 @@ func (node *Proxy) Search(ctx context.Context, request *milvuspb.SearchRequest)
optimizedSearch := true
resultSizeInsufficient := false
isTopkReduce := false
isRecallEvaluation := false
err2 := retry.Handle(ctx, func() (bool, error) {
rsp, resultSizeInsufficient, isTopkReduce, err = node.search(ctx, request, optimizedSearch)
if merr.Ok(rsp.GetStatus()) && resultSizeInsufficient && isTopkReduce && optimizedSearch && paramtable.Get().AutoIndexConfig.EnableResultLimitCheck.GetAsBool() {
rsp, resultSizeInsufficient, isTopkReduce, isRecallEvaluation, err = node.search(ctx, request, optimizedSearch, false)
if merr.Ok(rsp.GetStatus()) && optimizedSearch && resultSizeInsufficient && isTopkReduce && paramtable.Get().AutoIndexConfig.EnableResultLimitCheck.GetAsBool() {
// without optimize search
optimizedSearch = false
rsp, resultSizeInsufficient, isTopkReduce, err = node.search(ctx, request, optimizedSearch)
rsp, resultSizeInsufficient, isTopkReduce, isRecallEvaluation, err = node.search(ctx, request, optimizedSearch, false)
metrics.ProxyRetrySearchCount.WithLabelValues(
strconv.FormatInt(paramtable.GetNodeID(), 10),
metrics.SearchLabel,
Expand All @@ -2990,6 +2991,23 @@ func (node *Proxy) Search(ctx context.Context, request *milvuspb.SearchRequest)
if errors.Is(merr.Error(rsp.GetStatus()), merr.ErrInconsistentRequery) {
return true, merr.Error(rsp.GetStatus())
}
// search for ground truth and compute recall
if isRecallEvaluation && merr.Ok(rsp.GetStatus()) {
var rspGT *milvuspb.SearchResults
rspGT, _, _, _, err = node.search(ctx, request, false, true)
metrics.ProxyRecallSearchCount.WithLabelValues(
strconv.FormatInt(paramtable.GetNodeID(), 10),
metrics.SearchLabel,
request.GetCollectionName(),
).Inc()
if merr.Ok(rspGT.GetStatus()) {
return false, computeRecall(rsp.GetResults(), rspGT.GetResults())
}
if errors.Is(merr.Error(rspGT.GetStatus()), merr.ErrInconsistentRequery) {
return true, merr.Error(rspGT.GetStatus())
}
return false, merr.Error(rspGT.GetStatus())
}
return false, nil
})
if err2 != nil {
Expand All @@ -2998,7 +3016,7 @@ func (node *Proxy) Search(ctx context.Context, request *milvuspb.SearchRequest)
return rsp, err
}

func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest, optimizedSearch bool) (*milvuspb.SearchResults, bool, bool, error) {
func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest, optimizedSearch bool, isRecallEvaluation bool) (*milvuspb.SearchResults, bool, bool, bool, error) {
metrics.GetStats(ctx).
SetNodeID(paramtable.GetNodeID()).
SetInboundLabel(metrics.SearchLabel).
Expand All @@ -3013,7 +3031,7 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,
if err := merr.CheckHealthy(node.GetStateCode()); err != nil {
return &milvuspb.SearchResults{
Status: merr.Status(err),
}, false, false, nil
}, false, false, false, nil
}

method := "Search"
Expand All @@ -3034,7 +3052,7 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,
if err != nil {
return &milvuspb.SearchResults{
Status: merr.Status(err),
}, false, false, nil
}, false, false, false, nil
}

request.PlaceholderGroup = placeholderGroupBytes
Expand All @@ -3048,8 +3066,9 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,
commonpbutil.WithMsgType(commonpb.MsgType_Search),
commonpbutil.WithSourceID(paramtable.GetNodeID()),
),
ReqID: paramtable.GetNodeID(),
IsTopkReduce: optimizedSearch,
ReqID: paramtable.GetNodeID(),
IsTopkReduce: optimizedSearch,
IsRecallEvaluation: isRecallEvaluation,
},
request: request,
tr: timerecord.NewTimeRecorder("search"),
Expand Down Expand Up @@ -3103,7 +3122,7 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,

return &milvuspb.SearchResults{
Status: merr.Status(err),
}, false, false, nil
}, false, false, false, nil
}
tr.CtxRecord(ctx, "search request enqueue")

Expand All @@ -3129,7 +3148,7 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,

return &milvuspb.SearchResults{
Status: merr.Status(err),
}, false, false, nil
}, false, false, false, nil
}

span := tr.CtxRecord(ctx, "wait search result")
Expand Down Expand Up @@ -3186,7 +3205,7 @@ func (node *Proxy) search(ctx context.Context, request *milvuspb.SearchRequest,
metrics.ProxyReportValue.WithLabelValues(nodeID, hookutil.OpTypeSearch, dbName, username).Add(float64(v))
}
}
return qt.result, qt.resultSizeInsufficient, qt.isTopkReduce, nil
return qt.result, qt.resultSizeInsufficient, qt.isTopkReduce, qt.isRecallEvaluation, nil
}

func (node *Proxy) HybridSearch(ctx context.Context, request *milvuspb.HybridSearchRequest) (*milvuspb.SearchResults, error) {
Expand Down
6 changes: 6 additions & 0 deletions internal/proxy/task_search.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -63,6 +63,7 @@ type searchTask struct {
mustUsePartitionKey bool
resultSizeInsufficient bool
isTopkReduce bool
isRecallEvaluation bool

userOutputFields []string
userDynamicFields []string
Expand Down Expand Up @@ -621,10 +622,14 @@ func (t *searchTask) PostExecute(ctx context.Context) error {
t.queryChannelsTs = make(map[string]uint64)
t.relatedDataSize = 0
isTopkReduce := false
isRecallEvaluation := false
for _, r := range toReduceResults {
if r.GetIsTopkReduce() {
isTopkReduce = true
}
if r.GetIsRecallEvaluation() {
isRecallEvaluation = true
}
t.relatedDataSize += r.GetCostAggregation().GetTotalRelatedDataSize()
for ch, ts := range r.GetChannelsMvcc() {
t.queryChannelsTs[ch] = ts
Expand Down Expand Up @@ -703,6 +708,7 @@ func (t *searchTask) PostExecute(ctx context.Context) error {
}
t.resultSizeInsufficient = resultSizeInsufficient
t.isTopkReduce = isTopkReduce
t.isRecallEvaluation = isRecallEvaluation
t.result.CollectionName = t.collectionName
t.fillInFieldInfo()

Expand Down
69 changes: 69 additions & 0 deletions internal/proxy/util.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1079,6 +1079,75 @@ func translatePkOutputFields(schema *schemapb.CollectionSchema) ([]string, []int
return pkNames, fieldIDs
}

func recallCal[T string | int64](results []T, gts []T) float32 {
hit := 0
total := 0
for _, r := range results {
total++
for _, gt := range gts {
if r == gt {
hit++
break
}
}
}
return float32(hit) / float32(total)
}

func computeRecall(results *schemapb.SearchResultData, gts *schemapb.SearchResultData) error {
if results.GetNumQueries() != gts.GetNumQueries() {
return fmt.Errorf("num of queries is inconsistent between search results(%d) and ground truth(%d)", results.GetNumQueries(), gts.GetNumQueries())
}

switch results.GetIds().GetIdField().(type) {
case *schemapb.IDs_IntId:
switch gts.GetIds().GetIdField().(type) {
case *schemapb.IDs_IntId:
currentResultIndex := int64(0)
currentGTIndex := int64(0)
recalls := make([]float32, 0, results.GetNumQueries())
for i := 0; i < int(results.GetNumQueries()); i++ {
currentResultTopk := results.GetTopks()[i]
currentGTTopk := gts.GetTopks()[i]
recalls = append(recalls, recallCal(results.GetIds().GetIntId().GetData()[currentResultIndex:currentResultIndex+currentResultTopk],
gts.GetIds().GetIntId().GetData()[currentGTIndex:currentGTIndex+currentGTTopk]))
currentResultIndex += currentResultTopk
currentGTIndex += currentGTTopk
}
results.Recalls = recalls
return nil
case *schemapb.IDs_StrId:
return fmt.Errorf("pk type is inconsistent between search results(int64) and ground truth(string)")
default:
return fmt.Errorf("unsupported pk type")
}

case *schemapb.IDs_StrId:
switch gts.GetIds().GetIdField().(type) {
case *schemapb.IDs_StrId:
currentResultIndex := int64(0)
currentGTIndex := int64(0)
recalls := make([]float32, 0, results.GetNumQueries())
for i := 0; i < int(results.GetNumQueries()); i++ {
currentResultTopk := results.GetTopks()[i]
currentGTTopk := gts.GetTopks()[i]
recalls = append(recalls, recallCal(results.GetIds().GetStrId().GetData()[currentResultIndex:currentResultIndex+currentResultTopk],
gts.GetIds().GetStrId().GetData()[currentGTIndex:currentGTIndex+currentGTTopk]))
currentResultIndex += currentResultTopk
currentGTIndex += currentGTTopk
}
results.Recalls = recalls
return nil
case *schemapb.IDs_IntId:
return fmt.Errorf("pk type is inconsistent between search results(string) and ground truth(int64)")
default:
return fmt.Errorf("unsupported pk type")
}
default:
return fmt.Errorf("unsupported pk type")
}
}

// Support wildcard in output fields:
//
// "*" - all fields
Expand Down
162 changes: 162 additions & 0 deletions internal/proxy/util_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2672,3 +2672,165 @@ func TestValidateLoadFieldsList(t *testing.T) {
})
}
}

func TestComputeRecall(t *testing.T) {
t.Run("normal case1", func(t *testing.T) {
result1 := &schemapb.SearchResultData{
NumQueries: 3,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_StrId{
StrId: &schemapb.StringArray{
Data: []string{"11", "9", "8", "5", "3", "1"},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.1},
Topks: []int64{2, 2, 2},
}

gt := &schemapb.SearchResultData{
NumQueries: 3,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_StrId{
StrId: &schemapb.StringArray{
Data: []string{"11", "10", "8", "5", "3", "1"},
},
},
},
Scores: []float32{1.1, 0.98, 0.8, 0.5, 0.3, 0.1},
Topks: []int64{2, 2, 2},
}

err := computeRecall(result1, gt)
assert.NoError(t, err)
assert.Equal(t, result1.Recalls[0], float32(0.5))
assert.Equal(t, result1.Recalls[1], float32(1.0))
assert.Equal(t, result1.Recalls[2], float32(1.0))
})

t.Run("normal case2", func(t *testing.T) {
result1 := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 8, 5, 3, 1, 34, 23, 22, 21},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

gt := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 6, 5, 4, 1, 34, 23, 22, 20},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

err := computeRecall(result1, gt)
assert.NoError(t, err)
assert.Equal(t, result1.Recalls[0], float32(0.6))
assert.Equal(t, result1.Recalls[1], float32(0.8))
})

t.Run("not match size", func(t *testing.T) {
result1 := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 8, 5, 3, 1, 34, 23, 22, 21},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

gt := &schemapb.SearchResultData{
NumQueries: 1,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 6, 5, 4},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3},
Topks: []int64{5},
}

err := computeRecall(result1, gt)
assert.Error(t, err)
})

t.Run("not match type1", func(t *testing.T) {
result1 := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 8, 5, 3, 1, 34, 23, 22, 21},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

gt := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_StrId{
StrId: &schemapb.StringArray{
Data: []string{"11", "10", "8", "5", "3", "1", "23", "22", "21", "20"},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

err := computeRecall(result1, gt)
assert.Error(t, err)
})

t.Run("not match type2", func(t *testing.T) {
result1 := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_StrId{
StrId: &schemapb.StringArray{
Data: []string{"11", "10", "8", "5", "3", "1", "23", "22", "21", "20"},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

gt := &schemapb.SearchResultData{
NumQueries: 2,
Ids: &schemapb.IDs{
IdField: &schemapb.IDs_IntId{
IntId: &schemapb.LongArray{
Data: []int64{11, 9, 8, 5, 3, 1, 34, 23, 22, 21},
},
},
},
Scores: []float32{1.1, 0.9, 0.8, 0.5, 0.3, 0.8, 0.7, 0.6, 0.5, 0.4},
Topks: []int64{5, 5},
}

err := computeRecall(result1, gt)
assert.Error(t, err)
})
}
Loading
Loading