Christina Theodoris
commited on
Commit
·
4364d31
1
Parent(s):
76a78a0
move V1 autoformatting to after validate_options
Browse files
geneformer/classifier.py
CHANGED
@@ -234,10 +234,6 @@ class Classifier:
|
|
234 |
self.token_dictionary_file = token_dictionary_file
|
235 |
self.nproc = nproc
|
236 |
self.ngpu = ngpu
|
237 |
-
|
238 |
-
if self.model_version == "V1":
|
239 |
-
from . import TOKEN_DICTIONARY_FILE_30M
|
240 |
-
self.token_dictionary_file = TOKEN_DICTIONARY_FILE_30M
|
241 |
|
242 |
if self.training_args is None:
|
243 |
logger.warning(
|
@@ -258,7 +254,10 @@ class Classifier:
|
|
258 |
] = self.cell_state_dict["states"]
|
259 |
|
260 |
# load token dictionary (Ensembl IDs:token)
|
261 |
-
if self.
|
|
|
|
|
|
|
262 |
self.token_dictionary_file = TOKEN_DICTIONARY_FILE
|
263 |
with open(self.token_dictionary_file, "rb") as f:
|
264 |
self.gene_token_dict = pickle.load(f)
|
|
|
234 |
self.token_dictionary_file = token_dictionary_file
|
235 |
self.nproc = nproc
|
236 |
self.ngpu = ngpu
|
|
|
|
|
|
|
|
|
237 |
|
238 |
if self.training_args is None:
|
239 |
logger.warning(
|
|
|
254 |
] = self.cell_state_dict["states"]
|
255 |
|
256 |
# load token dictionary (Ensembl IDs:token)
|
257 |
+
if self.model_version == "V1":
|
258 |
+
from . import TOKEN_DICTIONARY_FILE_30M
|
259 |
+
self.token_dictionary_file = TOKEN_DICTIONARY_FILE_30M
|
260 |
+
elif self.token_dictionary_file is None:
|
261 |
self.token_dictionary_file = TOKEN_DICTIONARY_FILE
|
262 |
with open(self.token_dictionary_file, "rb") as f:
|
263 |
self.gene_token_dict = pickle.load(f)
|
geneformer/emb_extractor.py
CHANGED
@@ -518,6 +518,8 @@ class EmbExtractor:
|
|
518 |
self.summary_stat = summary_stat
|
519 |
self.exact_summary_stat = None
|
520 |
|
|
|
|
|
521 |
if self.model_version == "V1":
|
522 |
from . import TOKEN_DICTIONARY_FILE_30M
|
523 |
self.token_dictionary_file = TOKEN_DICTIONARY_FILE_30M
|
@@ -527,8 +529,6 @@ class EmbExtractor:
|
|
527 |
"model_version selected as V1 so changing emb_mode from 'cls' to 'cell' as V1 models do not have a <cls> token."
|
528 |
)
|
529 |
|
530 |
-
self.validate_options()
|
531 |
-
|
532 |
# load token dictionary (Ensembl IDs:token)
|
533 |
if self.token_dictionary_file is None:
|
534 |
token_dictionary_file = TOKEN_DICTIONARY_FILE
|
|
|
518 |
self.summary_stat = summary_stat
|
519 |
self.exact_summary_stat = None
|
520 |
|
521 |
+
self.validate_options()
|
522 |
+
|
523 |
if self.model_version == "V1":
|
524 |
from . import TOKEN_DICTIONARY_FILE_30M
|
525 |
self.token_dictionary_file = TOKEN_DICTIONARY_FILE_30M
|
|
|
529 |
"model_version selected as V1 so changing emb_mode from 'cls' to 'cell' as V1 models do not have a <cls> token."
|
530 |
)
|
531 |
|
|
|
|
|
532 |
# load token dictionary (Ensembl IDs:token)
|
533 |
if self.token_dictionary_file is None:
|
534 |
token_dictionary_file = TOKEN_DICTIONARY_FILE
|
geneformer/in_silico_perturber.py
CHANGED
@@ -231,7 +231,9 @@ class InSilicoPerturber:
|
|
231 |
self.nproc = nproc
|
232 |
self.model_version = model_version
|
233 |
self.token_dictionary_file = token_dictionary_file
|
234 |
-
self.clear_mem_ncells = clear_mem_ncells
|
|
|
|
|
235 |
|
236 |
if self.model_version == "V1":
|
237 |
from . import TOKEN_DICTIONARY_FILE_30M
|
@@ -245,10 +247,8 @@ class InSilicoPerturber:
|
|
245 |
self.emb_mode = "cell_and_gene"
|
246 |
logger.warning(
|
247 |
"model_version selected as V1 so changing emb_mode from 'cls_and_gene' to 'cell_and_gene' as V1 models do not have a <cls> token."
|
248 |
-
)
|
249 |
-
|
250 |
-
self.validate_options()
|
251 |
-
|
252 |
# load token dictionary (Ensembl IDs:token)
|
253 |
if self.token_dictionary_file is None:
|
254 |
token_dictionary_file = TOKEN_DICTIONARY_FILE
|
|
|
231 |
self.nproc = nproc
|
232 |
self.model_version = model_version
|
233 |
self.token_dictionary_file = token_dictionary_file
|
234 |
+
self.clear_mem_ncells = clear_mem_ncells
|
235 |
+
|
236 |
+
self.validate_options()
|
237 |
|
238 |
if self.model_version == "V1":
|
239 |
from . import TOKEN_DICTIONARY_FILE_30M
|
|
|
247 |
self.emb_mode = "cell_and_gene"
|
248 |
logger.warning(
|
249 |
"model_version selected as V1 so changing emb_mode from 'cls_and_gene' to 'cell_and_gene' as V1 models do not have a <cls> token."
|
250 |
+
)
|
251 |
+
|
|
|
|
|
252 |
# load token dictionary (Ensembl IDs:token)
|
253 |
if self.token_dictionary_file is None:
|
254 |
token_dictionary_file = TOKEN_DICTIONARY_FILE
|