diff --git a/tools/submission/generate_final_report.py b/tools/submission/generate_final_report.py
index a39098b73..f02c1fa01 100644
--- a/tools/submission/generate_final_report.py
+++ b/tools/submission/generate_final_report.py
@@ -111,6 +111,7 @@ def main():
                        'Latency (ms)',
                        'Samples/s',
                        'Queries/s',
+                       'Tokens/s',
                        'millijoules',
                        'Watts',
                    ]]
diff --git a/tools/submission/submission_checker.py b/tools/submission/submission_checker.py
index 95ee7d03d..692ff3f01 100755
--- a/tools/submission/submission_checker.py
+++ b/tools/submission/submission_checker.py
@@ -962,12 +962,13 @@ def check_accuracy_dir(config, model, path, verbose):
     is_valid = False
     all_accuracy_valid = True
     acc = None
-    result_acc = None
+    result_acc = {}
     hash_val = None
     target = config.get_accuracy_target(model)
     acc_upper_limit = config.get_accuracy_upper_limit(model)
     patterns = []
     acc_targets = []
+    acc_types = []
     if acc_upper_limit is not None:
         acc_limits = []
         up_patterns = []
@@ -981,10 +982,11 @@ def check_accuracy_dir(config, model, path, verbose):
         acc_type, acc_target = target[i:i+2]
         patterns.append(ACC_PATTERN[acc_type])
         acc_targets.append(acc_target)
+        acc_types.append(acc_type)
     acc_seen = [False for _ in acc_targets]
     with open(os.path.join(path, "accuracy.txt"), "r", encoding="utf-8") as f:
         for line in f:
-            for i, (pattern, acc_target) in enumerate(zip(patterns, acc_targets)):
+            for i, (pattern, acc_target, acc_type) in enumerate(zip(patterns, acc_targets, acc_types)):
                 m = re.match(pattern, line)
                 if m:
                     acc = m.group(1)
@@ -997,8 +999,8 @@ def check_accuracy_dir(config, model, path, verbose):
                 elif acc is not None:
                     all_accuracy_valid = False
                     log.warning("%s accuracy not met: expected=%f, found=%s", path, acc_target, acc)
-                if i == 0 and acc:
-                    result_acc = acc
+                if acc:
+                    result_acc[acc_type] = acc
                 acc = None
             if acc_upper_limit is not None:
                 for i, (pattern, acc_limit) in enumerate(zip(up_patterns, acc_limits)):
@@ -1590,6 +1592,38 @@ def log_result(
         if system_json.get("sw_notes"):
             notes = notes + ". " if notes else ""
             notes = notes + system_json.get("sw_notes")
+        special_unit_dict = {
+            "gptj-99": {
+                "SingleStream": "Latency (ms)",
+                "MultiStream": "Latency (ms)",
+                "Offline": "Tokens/s",
+                "Server": "Tokens/s",
+            },
+            "gptj-99.9": {
+                "SingleStream": "Latency (ms)",
+                "MultiStream": "Latency (ms)",
+                "Offline": "Tokens/s",
+                "Server": "Tokens/s",
+            },
+            "llama2-70b-99" : {
+                "SingleStream": "Latency (ms)",
+                "MultiStream": "Latency (ms)",
+                "Offline": "Tokens/s",
+                "Server": "Tokens/s",
+            },
+            "llama2-70b-99.9" : {
+                "SingleStream": "Latency (ms)",
+                "MultiStream": "Latency (ms)",
+                "Offline": "Tokens/s",
+                "Server": "Tokens/s",
+            },
+            "mixtral-8x7b" : {
+                "SingleStream": "Latency (ms)",
+                "MultiStream": "Latency (ms)",
+                "Offline": "Tokens/s",
+                "Server": "Tokens/s",
+            }
+        }
         unit_dict = {
             "SingleStream": "Latency (ms)",
             "MultiStream": "Latency (ms)",
@@ -1602,7 +1636,7 @@ def log_result(
             "Offline": "Watts",
             "Server": "Watts",
         }
-        unit = unit_dict[scenario_fixed]
+        unit = special_unit_dict.get(model_name, unit_dict)[scenario_fixed]
         power_unit = power_unit_dict[scenario_fixed]
 
         csv.write(
@@ -2012,6 +2046,7 @@ def log_result(
                                 acc_path,
                                 debug or is_closed_or_network,
                             )
+                            acc = json.dumps(acc).replace(",", " ").replace('"', "").replace("{", "").replace("}", "")
                             if mlperf_model in REQUIRED_ACC_BENCHMARK:
                                 if config.version in REQUIRED_ACC_BENCHMARK[mlperf_model]:
                                     extra_files_pass, missing_files = check_extra_files(acc_path, REQUIRED_ACC_BENCHMARK[mlperf_model][config.version])