@@ -68,7 +68,7 @@ def build_eval_args(eval_config, tp_size):
68
68
69
69
70
70
def generate_report (tp_size , eval_config , report_data , report_template ,
71
- output_path , env_config ):
71
+ report_output , env_config ):
72
72
env = Environment (loader = FileSystemLoader ('.' ))
73
73
template = env .get_template (str (report_template ))
74
74
model_args = build_model_args (eval_config , tp_size )
@@ -90,12 +90,12 @@ def generate_report(tp_size, eval_config, report_data, report_template,
90
90
batch_size = "auto" ,
91
91
num_fewshot = eval_params ["num_fewshot" ],
92
92
rows = report_data ["rows" ])
93
- with open (output_path , 'w' , encoding = 'utf-8' ) as f :
93
+ with open (report_output , 'w' , encoding = 'utf-8' ) as f :
94
94
f .write (report_content )
95
95
96
96
97
97
def test_lm_eval_correctness_param (config_filename , tp_size , report_template ,
98
- output_path , env_config ):
98
+ report_output , env_config ):
99
99
eval_config = yaml .safe_load (config_filename .read_text (encoding = "utf-8" ))
100
100
eval_params = build_eval_args (eval_config , tp_size )
101
101
results = lm_eval .simple_evaluate (** eval_params )
@@ -123,5 +123,5 @@ def test_lm_eval_correctness_param(config_filename, tp_size, report_template,
123
123
',' , '_stderr,' , 1 )]
124
124
})
125
125
generate_report (tp_size , eval_config , report_data , report_template ,
126
- output_path , env_config )
126
+ report_output , env_config )
127
127
assert success
0 commit comments