Update evaluation/run_generation_vllm.py

#5
by future7 - opened
Files changed (1) hide show
  1. evaluation/run_generation_vllm.py +8 -2
evaluation/run_generation_vllm.py CHANGED
@@ -69,14 +69,20 @@ def main():
69
  ## This changes the GPU support to 8
70
  model_vllm = LLM(model_path, tensor_parallel_size=8)
71
 
 
72
  output_list = []
73
  for prompt in prompt_list:
74
  prompt = bos_token + prompt
75
- output = model_vllm.generate([prompt], sampling_params)[0]
 
 
 
 
 
76
  generated_text = output.outputs[0].text
77
  generated_text = generated_text.strip().replace("\n", " ")
78
 
79
- # print("generated_text:", generated_text)
80
  output_list.append(generated_text)
81
 
82
  print("writing to %s" % output_datapath)
 
69
  ## This changes the GPU support to 8
70
  model_vllm = LLM(model_path, tensor_parallel_size=8)
71
 
72
+ input_list = []
73
  output_list = []
74
  for prompt in prompt_list:
75
  prompt = bos_token + prompt
76
+ input_list.append(prompt)
77
+
78
+
79
+ outputs = model_vllm.generate(input_list, sampling_params)[0]
80
+
81
+ for output in outputs:
82
  generated_text = output.outputs[0].text
83
  generated_text = generated_text.strip().replace("\n", " ")
84
 
85
+ print("generated_text:", generated_text)
86
  output_list.append(generated_text)
87
 
88
  print("writing to %s" % output_datapath)