Use CUDA event API for benchmarking

llama3.cu

@@ -783,16 +783,6 @@ int sample_argmax(float *probabilities, int n) {
     return max_i;
 }
 
-// ----------------------------------------------------------------------------
-// utilities: time
-// ----------------------------------------------------------------------------
-long time_in_ms() {
-    // return time in milliseconds, for benchmarking the model speed
-    struct timespec time;
-    clock_gettime(CLOCK_REALTIME, &time);
-    return time.tv_sec * 1000 + time.tv_nsec / 1000000;
-}
-
 // ----------------------------------------------------------------------------
 // generation loop
 // ----------------------------------------------------------------------------
@@ -812,10 +802,14 @@ void generate(Transformer *transformer, Tokenizer *tokenizer, char *prompt, int
     }
 
     // start the main loop
-    long start = 0;  // used to time our code, only initialized after first iteration
+    cudaEvent_t start, stop;  // CUDA events for measuring performance
+    cudaEventCreate(&start);
+    cudaEventCreate(&stop);
+
     int next;        // will store the next token in the sequence
     int token = prompt_tokens[0]; // kick off with the first token in the prompt
     int pos = 0;     // position in the sequence
+    cudaEventRecord(start);
     while (pos < max_new_tokens - 1) {
         // forward the transformer to get logits for the next token
         float *logits = forward(transformer, token, pos);
@@ -837,19 +831,20 @@ void generate(Transformer *transformer, Tokenizer *tokenizer, char *prompt, int
         safe_printf(piece); // same as printf("%s", piece), but skips "unsafe" bytes
         fflush(stdout);
         token = next;
-
-        // init the timer here because the first iteration can be slower
-        if (start == 0) { start = time_in_ms(); }
     }
+    cudaEventRecord(stop);
     printf("\n");
 
     // report achieved tok/s (Token count is assumed to be pos+1 because BOS token must be included)
     if (pos > 1) {
-        long end = time_in_ms();
+        cudaEventSynchronize(stop);
+        float milliseconds = 0;
+        cudaEventElapsedTime(&milliseconds, start, stop);
         fprintf(stderr, "Token count: %d, elapsed: %fs, %d tokens/s\n",
-                pos + 1, (float) (end - start) / 1000, (int) ((pos - 1) / (double) (end - start) * 1000));
+                pos + 1, milliseconds / 1000, (int) ((pos - 1) / milliseconds * 1000));
     }
-
+    cudaEventDestroy(start);
+    cudaEventDestroy(stop);
     free(prompt_tokens);
 }