Translation
COMET
zouharvi commited on
Commit
0d7eb85
·
verified ·
1 Parent(s): c51e020

update example

Browse files
Files changed (1) hide show
  1. README.md +13 -9
README.md CHANGED
@@ -123,20 +123,24 @@ import comet_early_exit
123
  model = comet_early_exit.load_from_checkpoint(comet_early_exit.download_model("zouharvi/COMET-partial"))
124
  data = [
125
  {
126
- "src": "Will I receive my food in 10 to 15 minutes?",
127
- "mt": "Ich bekomme",
128
  },
129
  {
130
- "src": "Will I receive my food in 10 to 15 minutes?",
131
- "mt": "Bekomme ich",
132
  },
133
  {
134
- "src": "Will I receive my food in 10 to 15 minutes?",
135
- "mt": "Bekomme ich mein Essen in",
136
  },
137
  {
138
- "src": "Will I receive my food in 10 to 15 minutes?",
139
- "mt": "Bekomme ich mein Essen in 10 bis 15 Minuten?",
 
 
 
 
140
  }
141
  ]
142
  model_output = model.predict(data, batch_size=8, gpus=1)
@@ -144,7 +148,7 @@ print("scores", model_output["scores"])
144
  ```
145
  Outputs (formatted):
146
  ```
147
- scores 89.58 88.99 89.94 89.94
148
  ```
149
 
150
  This model is based on the work [Early-Exit and Instant Confidence Translation Quality Estimation](http://arxiv.org/abs/2502.14429) which can be cited as:
 
123
  model = comet_early_exit.load_from_checkpoint(comet_early_exit.download_model("zouharvi/COMET-partial"))
124
  data = [
125
  {
126
+ "src": "I want to receive my food in 10 to 15 minutes.",
127
+ "mt": "Ich werde",
128
  },
129
  {
130
+ "src": "I want to receive my food in 10 to 15 minutes.",
131
+ "mt": "Ich möchte",
132
  },
133
  {
134
+ "src": "I want to receive my food in 10 to 15 minutes.",
135
+ "mt": "Ich möchte mein Essen in",
136
  },
137
  {
138
+ "src": "I want to receive my food in 10 to 15 minutes.",
139
+ "mt": "Ich möchte mein Essen in 10 bis 15 Minuten erhalten.",
140
+ },
141
+ {
142
+ "src": "I want to receive my food in 10 to 15 minutes.",
143
+ "mt": "Ich möchte mein Essen in 10 bis 15 Minuten bekommen.",
144
  }
145
  ]
146
  model_output = model.predict(data, batch_size=8, gpus=1)
 
148
  ```
149
  Outputs (formatted):
150
  ```
151
+ scores 89.26 89.45 89.51 89.48 89.66
152
  ```
153
 
154
  This model is based on the work [Early-Exit and Instant Confidence Translation Quality Estimation](http://arxiv.org/abs/2502.14429) which can be cited as: