Skip to content

armgit / ruflo

Public Imported from GitHub
0

🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated

checkpoint: File edit:

rUv0afde59891aa
diff --git a/v3/@claude-flow/performance/src/attention-integration.ts b/v3/@claude-flow/performance/src/attention-integration.ts
index ae3551af38..2acc2c088b 100644
--- a/v3/@claude-flow/performance/src/attention-integration.ts
+++ b/v3/@claude-flow/performance/src/attention-integration.ts
@@ -98,7 +98,7 @@ export class FlashAttentionOptimizer {
* @param input - Query, keys, and values for attention computation
* @returns Optimized attention output with performance metrics
*/
- async optimize(input: AttentionInput): Promise<AttentionOutput> {
+ optimize(input: AttentionInput): AttentionOutput {
const startTime = performance.now();
const startMemory = this.getMemoryUsage();
@@ -107,8 +107,8 @@ export class FlashAttentionOptimizer {
const keys = input.keys.map(k => this.ensureFloat32Array(k));
const values = input.values.map(v => this.ensureFloat32Array(v));
- // Use async Flash Attention for best performance
- const result = await computeFlashAttentionAsync(query, keys, values);
+ // Use synchronous Flash Attention with raw Float32Arrays
+ const result = this.flashAttention.computeRaw(query, keys, values);
const executionTimeMs = performance.now() - startTime;
const endMemory = this.getMemoryUsage();