kimsan0622
commited on
Commit
•
865e4c1
1
Parent(s):
5a34584
Upload model
Browse files- config.json +6 -2
- modeling_veld.py +8 -1
config.json
CHANGED
@@ -15,6 +15,7 @@
|
|
15 |
"T5DualDecoderDoubleHeadsModel"
|
16 |
],
|
17 |
"bad_words_ids": null,
|
|
|
18 |
"bos_token_id": null,
|
19 |
"chunk_size_feed_forward": 0,
|
20 |
"cross_attention_hidden_size": null,
|
@@ -73,6 +74,7 @@
|
|
73 |
"return_dict": true,
|
74 |
"return_dict_in_generate": false,
|
75 |
"sep_token_id": null,
|
|
|
76 |
"task_specific_params": null,
|
77 |
"temperature": 1.0,
|
78 |
"tf_legacy_loss": false,
|
@@ -83,7 +85,7 @@
|
|
83 |
"top_p": 1.0,
|
84 |
"torch_dtype": null,
|
85 |
"torchscript": false,
|
86 |
-
"transformers_version": "4.
|
87 |
"typical_p": 1.0,
|
88 |
"use_bfloat16": false,
|
89 |
"use_cache": true,
|
@@ -97,6 +99,7 @@
|
|
97 |
],
|
98 |
"attention_probs_dropout_prob": 0.0,
|
99 |
"bad_words_ids": null,
|
|
|
100 |
"bos_token_id": null,
|
101 |
"chunk_size_feed_forward": 0,
|
102 |
"cross_attention_hidden_size": null,
|
@@ -153,6 +156,7 @@
|
|
153 |
"return_dict": true,
|
154 |
"return_dict_in_generate": false,
|
155 |
"sep_token_id": null,
|
|
|
156 |
"task_specific_params": null,
|
157 |
"temperature": 1.0,
|
158 |
"tf_legacy_loss": false,
|
@@ -163,7 +167,7 @@
|
|
163 |
"top_p": 1.0,
|
164 |
"torch_dtype": null,
|
165 |
"torchscript": false,
|
166 |
-
"transformers_version": "4.
|
167 |
"typical_p": 1.0,
|
168 |
"use_bfloat16": false
|
169 |
},
|
|
|
15 |
"T5DualDecoderDoubleHeadsModel"
|
16 |
],
|
17 |
"bad_words_ids": null,
|
18 |
+
"begin_suppress_tokens": null,
|
19 |
"bos_token_id": null,
|
20 |
"chunk_size_feed_forward": 0,
|
21 |
"cross_attention_hidden_size": null,
|
|
|
74 |
"return_dict": true,
|
75 |
"return_dict_in_generate": false,
|
76 |
"sep_token_id": null,
|
77 |
+
"suppress_tokens": null,
|
78 |
"task_specific_params": null,
|
79 |
"temperature": 1.0,
|
80 |
"tf_legacy_loss": false,
|
|
|
85 |
"top_p": 1.0,
|
86 |
"torch_dtype": null,
|
87 |
"torchscript": false,
|
88 |
+
"transformers_version": "4.24.0",
|
89 |
"typical_p": 1.0,
|
90 |
"use_bfloat16": false,
|
91 |
"use_cache": true,
|
|
|
99 |
],
|
100 |
"attention_probs_dropout_prob": 0.0,
|
101 |
"bad_words_ids": null,
|
102 |
+
"begin_suppress_tokens": null,
|
103 |
"bos_token_id": null,
|
104 |
"chunk_size_feed_forward": 0,
|
105 |
"cross_attention_hidden_size": null,
|
|
|
156 |
"return_dict": true,
|
157 |
"return_dict_in_generate": false,
|
158 |
"sep_token_id": null,
|
159 |
+
"suppress_tokens": null,
|
160 |
"task_specific_params": null,
|
161 |
"temperature": 1.0,
|
162 |
"tf_legacy_loss": false,
|
|
|
167 |
"top_p": 1.0,
|
168 |
"torch_dtype": null,
|
169 |
"torchscript": false,
|
170 |
+
"transformers_version": "4.24.0",
|
171 |
"typical_p": 1.0,
|
172 |
"use_bfloat16": false
|
173 |
},
|
modeling_veld.py
CHANGED
@@ -1886,7 +1886,7 @@ class VELDModel(PreTrainedModel):
|
|
1886 |
|
1887 |
encoder_hidden_states = None if encoder_outputs is None else encoder_outputs[0]
|
1888 |
pooler_output_local = None if encoder_outputs is None else self.local_pooling(encoder_hidden_states)
|
1889 |
-
pooler_output_global = None if encoder_outputs is None else self.global_pooling(pooler_output_local).squeeze(1)
|
1890 |
|
1891 |
# optionally project encoder_hidden_states
|
1892 |
if (
|
@@ -2083,6 +2083,13 @@ if __name__ == "__main__":
|
|
2083 |
print(outputs.loss)
|
2084 |
print(outputs.c_loss)
|
2085 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2086 |
# print(outputs)
|
2087 |
|
2088 |
# outputs = model.generate(
|
|
|
1886 |
|
1887 |
encoder_hidden_states = None if encoder_outputs is None else encoder_outputs[0]
|
1888 |
pooler_output_local = None if encoder_outputs is None else self.local_pooling(encoder_hidden_states)
|
1889 |
+
pooler_output_global = None if encoder_outputs is None or return_contrastive_loss is None else self.global_pooling(pooler_output_local).squeeze(1)
|
1890 |
|
1891 |
# optionally project encoder_hidden_states
|
1892 |
if (
|
|
|
2083 |
print(outputs.loss)
|
2084 |
print(outputs.c_loss)
|
2085 |
|
2086 |
+
outputs = model(
|
2087 |
+
pixel_values=pixel_values,
|
2088 |
+
labels=inps.input_ids,
|
2089 |
+
decoder_attention_mask=inps.attention_mask)
|
2090 |
+
print(outputs.loss)
|
2091 |
+
print(outputs.c_loss)
|
2092 |
+
|
2093 |
# print(outputs)
|
2094 |
|
2095 |
# outputs = model.generate(
|