Upload model.py with huggingface_hub
Browse files
model.py
CHANGED
|
@@ -281,6 +281,8 @@ class BT4Config(PretrainedConfig):
|
|
| 281 |
self.smol_activation = smol_activation
|
| 282 |
|
| 283 |
class BT4(PreTrainedModel):
|
|
|
|
|
|
|
| 284 |
def __init__(self, config=None, embedding_size=1024, embedding_dense_sz=512, encoder_layers=15, encoder_d_model=1024, encoder_heads=32, encoder_dff=1536, dropout_rate=0.0, pol_embedding_size=1024, policy_d_model=1024, val_embedding_size=128, default_activation=Mish(),
|
| 285 |
use_smolgen=True, smol_hidden_channels=32, smol_hidden_sz=256, smol_gen_sz=256, smol_activation='swish'):
|
| 286 |
# Initialize PreTrainedModel with config
|
|
|
|
| 281 |
self.smol_activation = smol_activation
|
| 282 |
|
| 283 |
class BT4(PreTrainedModel):
|
| 284 |
+
config_class = BT4Config
|
| 285 |
+
|
| 286 |
def __init__(self, config=None, embedding_size=1024, embedding_dense_sz=512, encoder_layers=15, encoder_d_model=1024, encoder_heads=32, encoder_dff=1536, dropout_rate=0.0, pol_embedding_size=1024, policy_d_model=1024, val_embedding_size=128, default_activation=Mish(),
|
| 287 |
use_smolgen=True, smol_hidden_channels=32, smol_hidden_sz=256, smol_gen_sz=256, smol_activation='swish'):
|
| 288 |
# Initialize PreTrainedModel with config
|