@@ -156,16 +156,15 @@ def _batched_encode(
156156 and len (batch ) < batch_size
157157 and batch_tokens < self ._max_tokens_per_batch
158158 ):
159- n_tokens = len (
160- self ._client .tokenize ([sentences [index ]], model = self ._model_name )[0 ]
161- )
159+ txt = sentences [index ] if len (sentences [index ]) > 0 else " "
160+ n_tokens = len (self ._client .tokenize ([txt ], model = self ._model_name )[0 ])
162161 if (
163162 batch_tokens + n_tokens > self ._max_tokens_per_batch
164163 and len (batch ) > 0
165164 ):
166165 break
167166 batch_tokens += n_tokens
168- batch .append (sentences [ index ] )
167+ batch .append (txt )
169168 index += 1
170169
171170 embeddings .extend (
@@ -249,7 +248,7 @@ def _batched_encode(
249248 n_parameters = None ,
250249 memory_usage_mb = None ,
251250 license = None ,
252- reference = "https://docs .voyageai.com/docs/embeddings " ,
251+ reference = "https://blog .voyageai.com/2025/05/20/voyage-3-5/ " ,
253252 similarity_fn_name = "cosine" ,
254253 framework = ["API" ],
255254 use_instructions = True ,
@@ -274,7 +273,7 @@ def _batched_encode(
274273 n_parameters = None ,
275274 memory_usage_mb = None ,
276275 license = None ,
277- reference = "https://docs .voyageai.com/docs/flexible-dimensions-and-quantization " ,
276+ reference = "https://blog .voyageai.com/2025/05/20/voyage-3-5/ " ,
278277 similarity_fn_name = "cosine" ,
279278 framework = ["API" ],
280279 use_instructions = True ,
@@ -300,7 +299,7 @@ def _batched_encode(
300299 n_parameters = None ,
301300 memory_usage_mb = None ,
302301 license = None ,
303- reference = "https://docs .voyageai.com/docs/flexible-dimensions-and-quantization " ,
302+ reference = "https://blog .voyageai.com/2025/05/20/voyage-3-5/ " ,
304303 similarity_fn_name = "cosine" ,
305304 framework = ["API" ],
306305 use_instructions = True ,
0 commit comments