SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

[Flax] Fixes typo in Bart based Flax Models (#13565)

Browse Source
This commit is contained in:
Bhadresh Savani
2021-09-15 11:03:52 +05:30
committed by GitHub
parent 7bd16b8776
commit 3fbb55c757
5 changed files with 20 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/transformers/models/bart/modeling_flax_bart.py
View File

@@ -406,7 +406,7 @@ class FlaxBartEncoderLayer(nn.Module):
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.fc1 = nn.Dense( self.fc1 = nn.Dense(
self.config.encoder_ffn_dim, self.config.encoder_ffn_dim,
dtype=self.dtype, dtype=self.dtype,
@@ -433,7 +433,7 @@ class FlaxBartEncoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states
@@ -515,7 +515,7 @@ class FlaxBartDecoderLayer(nn.Module):
) )
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.encoder_attn = FlaxBartAttention( self.encoder_attn = FlaxBartAttention(
@@ -572,7 +572,7 @@ class FlaxBartDecoderLayer(nn.Module):
# Fully Connected # Fully Connected
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states

8
src/transformers/models/marian/modeling_flax_marian.py
View File

@@ -411,7 +411,7 @@ class FlaxMarianEncoderLayer(nn.Module):
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.fc1 = nn.Dense( self.fc1 = nn.Dense(
self.config.encoder_ffn_dim, self.config.encoder_ffn_dim,
dtype=self.dtype, dtype=self.dtype,
@@ -438,7 +438,7 @@ class FlaxMarianEncoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states
@@ -523,7 +523,7 @@ class FlaxMarianDecoderLayer(nn.Module):
) )
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.encoder_attn = FlaxMarianAttention( self.encoder_attn = FlaxMarianAttention(
@@ -580,7 +580,7 @@ class FlaxMarianDecoderLayer(nn.Module):
# Fully Connected # Fully Connected
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states

8
src/transformers/models/mbart/modeling_flax_mbart.py
View File

@@ -417,7 +417,7 @@ class FlaxMBartEncoderLayer(nn.Module):
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.fc1 = nn.Dense( self.fc1 = nn.Dense(
self.config.encoder_ffn_dim, self.config.encoder_ffn_dim,
dtype=self.dtype, dtype=self.dtype,
@@ -444,7 +444,7 @@ class FlaxMBartEncoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.final_layer_norm(hidden_states) hidden_states = self.final_layer_norm(hidden_states)
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states
@@ -527,7 +527,7 @@ class FlaxMBartDecoderLayer(nn.Module):
) )
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.encoder_attn = FlaxMBartAttention( self.encoder_attn = FlaxMBartAttention(
@@ -585,7 +585,7 @@ class FlaxMBartDecoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.final_layer_norm(hidden_states) hidden_states = self.final_layer_norm(hidden_states)
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states

8
src/transformers/models/pegasus/modeling_flax_pegasus.py
View File

@@ -411,7 +411,7 @@ class FlaxPegasusEncoderLayer(nn.Module):
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.fc1 = nn.Dense( self.fc1 = nn.Dense(
self.config.encoder_ffn_dim, self.config.encoder_ffn_dim,
dtype=self.dtype, dtype=self.dtype,
@@ -438,7 +438,7 @@ class FlaxPegasusEncoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.final_layer_norm(hidden_states) hidden_states = self.final_layer_norm(hidden_states)
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states
@@ -522,7 +522,7 @@ class FlaxPegasusDecoderLayer(nn.Module):
) )
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.encoder_attn = FlaxPegasusAttention( self.encoder_attn = FlaxPegasusAttention(
@@ -580,7 +580,7 @@ class FlaxPegasusDecoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.final_layer_norm(hidden_states) hidden_states = self.final_layer_norm(hidden_states)
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states

8
templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/modeling_flax_{{cookiecutter.lowercase_modelname}}.py
View File

@@ -1432,7 +1432,7 @@ class Flax{{cookiecutter.camelcase_modelname}}EncoderLayer(nn.Module):
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.fc1 = nn.Dense( self.fc1 = nn.Dense(
self.config.encoder_ffn_dim, self.config.encoder_ffn_dim,
dtype=self.dtype, dtype=self.dtype,
@@ -1459,7 +1459,7 @@ class Flax{{cookiecutter.camelcase_modelname}}EncoderLayer(nn.Module):
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states
@@ -1541,7 +1541,7 @@ class Flax{{cookiecutter.camelcase_modelname}}DecoderLayer(nn.Module):
) )
self.dropout_layer = nn.Dropout(rate=self.config.dropout) self.dropout_layer = nn.Dropout(rate=self.config.dropout)
self.activation_fn = ACT2FN[self.config.activation_function] self.activation_fn = ACT2FN[self.config.activation_function]
self.acticvation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout) self.activation_dropout_layer = nn.Dropout(rate=self.config.activation_dropout)
self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype) self.self_attn_layer_norm = nn.LayerNorm(dtype=self.dtype)
self.encoder_attn = Flax{{cookiecutter.camelcase_modelname}}Attention( self.encoder_attn = Flax{{cookiecutter.camelcase_modelname}}Attention(
@@ -1598,7 +1598,7 @@ class Flax{{cookiecutter.camelcase_modelname}}DecoderLayer(nn.Module):
# Fully Connected # Fully Connected
residual = hidden_states residual = hidden_states
hidden_states = self.activation_fn(self.fc1(hidden_states)) hidden_states = self.activation_fn(self.fc1(hidden_states))
hidden_states = self.acticvation_dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.activation_dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = self.fc2(hidden_states) hidden_states = self.fc2(hidden_states)
hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic) hidden_states = self.dropout_layer(hidden_states, deterministic=deterministic)
hidden_states = residual + hidden_states hidden_states = residual + hidden_states