ydshieh
commited on
Commit
•
87485e5
1
Parent(s):
3a17811
final ver
Browse files- vit_gpt2/modeling_flax_gpt2.py +159 -37
vit_gpt2/modeling_flax_gpt2.py
CHANGED
@@ -24,7 +24,7 @@ from flax.linen.attention import dot_product_attention_weights
|
|
24 |
from jax import lax
|
25 |
|
26 |
from ...file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
|
27 |
-
from ...modeling_flax_outputs import
|
28 |
from ...modeling_flax_utils import ACT2FN, FlaxPreTrainedModel, append_call_sample_docstring
|
29 |
from ...utils import logging
|
30 |
from .configuration_gpt2 import GPT2Config
|
@@ -117,6 +117,8 @@ class FlaxConv1D(nn.Module):
|
|
117 |
class FlaxGPT2Attention(nn.Module):
|
118 |
config: GPT2Config
|
119 |
dtype: jnp.dtype = jnp.float32
|
|
|
|
|
120 |
|
121 |
def setup(self):
|
122 |
config = self.config
|
@@ -124,10 +126,19 @@ class FlaxGPT2Attention(nn.Module):
|
|
124 |
self.num_heads = config.num_attention_heads
|
125 |
self.head_dim = self.embed_dim // self.num_heads
|
126 |
|
127 |
-
|
|
|
|
|
|
|
|
|
128 |
self.c_proj = FlaxConv1D(self.embed_dim, dtype=self.dtype)
|
|
|
129 |
self.resid_dropout = nn.Dropout(rate=config.resid_pdrop)
|
130 |
-
|
|
|
|
|
|
|
|
|
131 |
|
132 |
def _split_heads(self, hidden_states):
|
133 |
return hidden_states.reshape(hidden_states.shape[:2] + (self.num_heads, self.head_dim))
|
@@ -170,13 +181,26 @@ class FlaxGPT2Attention(nn.Module):
|
|
170 |
def __call__(
|
171 |
self,
|
172 |
hidden_states,
|
|
|
173 |
attention_mask=None,
|
174 |
deterministic: bool = True,
|
175 |
init_cache: bool = False,
|
176 |
output_attentions: bool = False,
|
177 |
):
|
178 |
-
|
179 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
180 |
|
181 |
query = self._split_heads(query)
|
182 |
key = self._split_heads(key)
|
@@ -184,20 +208,25 @@ class FlaxGPT2Attention(nn.Module):
|
|
184 |
|
185 |
query_length, key_length = query.shape[1], key.shape[1]
|
186 |
|
187 |
-
if self.
|
188 |
-
|
189 |
-
|
190 |
-
|
191 |
-
|
192 |
-
|
193 |
-
|
194 |
-
|
195 |
-
|
196 |
-
|
197 |
-
|
198 |
-
|
199 |
-
attention_mask
|
200 |
-
|
|
|
|
|
|
|
|
|
|
|
201 |
|
202 |
dropout_rng = None
|
203 |
if not deterministic and self.config.attn_pdrop > 0.0:
|
@@ -205,15 +234,18 @@ class FlaxGPT2Attention(nn.Module):
|
|
205 |
|
206 |
# During fast autoregressive decoding, we feed one position at a time,
|
207 |
# and cache the keys and values step by step.
|
208 |
-
if self.has_variable("cache", "cached_key") or init_cache:
|
209 |
key, value, attention_mask = self._concatenate_to_cache(key, value, query, attention_mask)
|
210 |
|
211 |
# transform boolean mask into float mask
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
-
|
|
|
|
|
|
|
217 |
|
218 |
# usual dot product attention
|
219 |
attn_weights = dot_product_attention_weights(
|
@@ -267,19 +299,31 @@ class FlaxGPT2Block(nn.Module):
|
|
267 |
self.ln_1 = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
268 |
self.attn = FlaxGPT2Attention(self.config, dtype=self.dtype)
|
269 |
self.ln_2 = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
270 |
self.mlp = FlaxGPT2MLP(self.config, inner_dim, dtype=self.dtype)
|
271 |
|
272 |
def __call__(
|
273 |
self,
|
274 |
hidden_states,
|
275 |
attention_mask=None,
|
|
|
|
|
276 |
deterministic: bool = True,
|
277 |
init_cache: bool = False,
|
278 |
output_attentions: bool = False,
|
279 |
):
|
280 |
residual = hidden_states
|
281 |
hidden_states = self.ln_1(hidden_states)
|
282 |
-
|
283 |
hidden_states,
|
284 |
attention_mask=attention_mask,
|
285 |
deterministic=deterministic,
|
@@ -287,16 +331,53 @@ class FlaxGPT2Block(nn.Module):
|
|
287 |
output_attentions=output_attentions,
|
288 |
)
|
289 |
# residual connection
|
290 |
-
attn_output =
|
|
|
|
|
291 |
hidden_states = attn_output + residual
|
292 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
293 |
residual = hidden_states
|
294 |
hidden_states = self.ln_2(hidden_states)
|
295 |
feed_forward_hidden_states = self.mlp(hidden_states, deterministic=deterministic)
|
296 |
# residual connection
|
297 |
hidden_states = residual + feed_forward_hidden_states
|
298 |
|
299 |
-
|
|
|
|
|
300 |
|
301 |
|
302 |
class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
@@ -328,7 +409,19 @@ class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
|
328 |
params_rng, dropout_rng = jax.random.split(rng)
|
329 |
rngs = {"params": params_rng, "dropout": dropout_rng}
|
330 |
|
331 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
332 |
|
333 |
def init_cache(self, batch_size, max_length):
|
334 |
r"""
|
@@ -355,6 +448,8 @@ class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
|
355 |
input_ids,
|
356 |
attention_mask=None,
|
357 |
position_ids=None,
|
|
|
|
|
358 |
params: dict = None,
|
359 |
past_key_values: dict = None,
|
360 |
dropout_rng: jax.random.PRNGKey = None,
|
@@ -369,6 +464,10 @@ class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
|
369 |
)
|
370 |
return_dict = return_dict if return_dict is not None else self.config.return_dict
|
371 |
|
|
|
|
|
|
|
|
|
372 |
batch_size, sequence_length = input_ids.shape
|
373 |
|
374 |
if position_ids is None:
|
@@ -399,6 +498,8 @@ class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
|
399 |
jnp.array(input_ids, dtype="i4"),
|
400 |
jnp.array(attention_mask, dtype="i4"),
|
401 |
jnp.array(position_ids, dtype="i4"),
|
|
|
|
|
402 |
not train,
|
403 |
False,
|
404 |
output_attentions,
|
@@ -433,6 +534,8 @@ class FlaxGPT2BlockCollection(nn.Module):
|
|
433 |
self,
|
434 |
hidden_states,
|
435 |
attention_mask=None,
|
|
|
|
|
436 |
deterministic: bool = True,
|
437 |
init_cache: bool = False,
|
438 |
output_attentions: bool = False,
|
@@ -441,6 +544,7 @@ class FlaxGPT2BlockCollection(nn.Module):
|
|
441 |
):
|
442 |
all_attentions = () if output_attentions else None
|
443 |
all_hidden_states = () if output_hidden_states else None
|
|
|
444 |
|
445 |
for block in self.blocks:
|
446 |
if output_hidden_states:
|
@@ -449,6 +553,8 @@ class FlaxGPT2BlockCollection(nn.Module):
|
|
449 |
layer_outputs = block(
|
450 |
hidden_states,
|
451 |
attention_mask,
|
|
|
|
|
452 |
deterministic=deterministic,
|
453 |
init_cache=init_cache,
|
454 |
output_attentions=output_attentions,
|
@@ -458,19 +564,22 @@ class FlaxGPT2BlockCollection(nn.Module):
|
|
458 |
if output_attentions:
|
459 |
all_attentions += (layer_outputs[1],)
|
460 |
|
|
|
|
|
|
|
461 |
if output_hidden_states:
|
462 |
all_hidden_states += (hidden_states,)
|
463 |
|
464 |
-
outputs =
|
465 |
|
466 |
if not return_dict:
|
467 |
return tuple(v for v in outputs if v is not None)
|
468 |
|
469 |
-
return
|
470 |
last_hidden_state=hidden_states,
|
471 |
-
past_key_values=None,
|
472 |
hidden_states=all_hidden_states,
|
473 |
attentions=all_attentions,
|
|
|
474 |
)
|
475 |
|
476 |
|
@@ -502,6 +611,8 @@ class FlaxGPT2Module(nn.Module):
|
|
502 |
input_ids,
|
503 |
attention_mask,
|
504 |
position_ids,
|
|
|
|
|
505 |
deterministic=True,
|
506 |
init_cache: bool = False,
|
507 |
output_attentions: bool = False,
|
@@ -517,6 +628,8 @@ class FlaxGPT2Module(nn.Module):
|
|
517 |
outputs = self.h(
|
518 |
hidden_states,
|
519 |
attention_mask,
|
|
|
|
|
520 |
deterministic=deterministic,
|
521 |
init_cache=init_cache,
|
522 |
output_attentions=output_attentions,
|
@@ -530,10 +643,11 @@ class FlaxGPT2Module(nn.Module):
|
|
530 |
if not return_dict:
|
531 |
return (hidden_states,) + outputs[1:]
|
532 |
|
533 |
-
return
|
534 |
last_hidden_state=hidden_states,
|
535 |
hidden_states=outputs.hidden_states,
|
536 |
attentions=outputs.attentions,
|
|
|
537 |
)
|
538 |
|
539 |
|
@@ -546,7 +660,7 @@ class FlaxGPT2Model(FlaxGPT2PreTrainedModel):
|
|
546 |
|
547 |
|
548 |
append_call_sample_docstring(
|
549 |
-
FlaxGPT2Model, _TOKENIZER_FOR_DOC, _CHECKPOINT_FOR_DOC,
|
550 |
)
|
551 |
|
552 |
|
@@ -568,6 +682,8 @@ class FlaxGPT2LMHeadModule(nn.Module):
|
|
568 |
input_ids,
|
569 |
attention_mask,
|
570 |
position_ids,
|
|
|
|
|
571 |
deterministic: bool = True,
|
572 |
init_cache: bool = False,
|
573 |
output_attentions: bool = False,
|
@@ -578,6 +694,8 @@ class FlaxGPT2LMHeadModule(nn.Module):
|
|
578 |
input_ids,
|
579 |
attention_mask,
|
580 |
position_ids,
|
|
|
|
|
581 |
deterministic=deterministic,
|
582 |
init_cache=init_cache,
|
583 |
output_attentions=output_attentions,
|
@@ -596,8 +714,12 @@ class FlaxGPT2LMHeadModule(nn.Module):
|
|
596 |
if not return_dict:
|
597 |
return (lm_logits,) + outputs[1:]
|
598 |
|
599 |
-
return
|
600 |
-
|
|
|
|
|
|
|
|
|
601 |
|
602 |
@add_start_docstrings(
|
603 |
"""
|
@@ -637,5 +759,5 @@ class FlaxGPT2LMHeadModel(FlaxGPT2PreTrainedModel):
|
|
637 |
|
638 |
|
639 |
append_call_sample_docstring(
|
640 |
-
FlaxGPT2LMHeadModel, _TOKENIZER_FOR_DOC, _CHECKPOINT_FOR_DOC,
|
641 |
)
|
|
|
24 |
from jax import lax
|
25 |
|
26 |
from ...file_utils import add_start_docstrings, add_start_docstrings_to_model_forward
|
27 |
+
from ...modeling_flax_outputs import FlaxBaseModelOutputWithPastAndCrossAttentions, FlaxCausalLMOutputWithCrossAttentions
|
28 |
from ...modeling_flax_utils import ACT2FN, FlaxPreTrainedModel, append_call_sample_docstring
|
29 |
from ...utils import logging
|
30 |
from .configuration_gpt2 import GPT2Config
|
|
|
117 |
class FlaxGPT2Attention(nn.Module):
|
118 |
config: GPT2Config
|
119 |
dtype: jnp.dtype = jnp.float32
|
120 |
+
causal: bool = True
|
121 |
+
is_cross_attention: bool = False
|
122 |
|
123 |
def setup(self):
|
124 |
config = self.config
|
|
|
126 |
self.num_heads = config.num_attention_heads
|
127 |
self.head_dim = self.embed_dim // self.num_heads
|
128 |
|
129 |
+
if self.is_cross_attention:
|
130 |
+
self.c_attn = FlaxConv1D(2 * self.embed_dim, dtype=self.dtype)
|
131 |
+
self.q_attn = FlaxConv1D(self.embed_dim, dtype=self.dtype)
|
132 |
+
else:
|
133 |
+
self.c_attn = FlaxConv1D(3 * self.embed_dim, dtype=self.dtype)
|
134 |
self.c_proj = FlaxConv1D(self.embed_dim, dtype=self.dtype)
|
135 |
+
|
136 |
self.resid_dropout = nn.Dropout(rate=config.resid_pdrop)
|
137 |
+
|
138 |
+
if self.causal:
|
139 |
+
self.causal_mask = make_causal_mask(
|
140 |
+
jnp.ones((1, config.max_position_embeddings), dtype="bool"), dtype="bool"
|
141 |
+
)
|
142 |
|
143 |
def _split_heads(self, hidden_states):
|
144 |
return hidden_states.reshape(hidden_states.shape[:2] + (self.num_heads, self.head_dim))
|
|
|
181 |
def __call__(
|
182 |
self,
|
183 |
hidden_states,
|
184 |
+
key_value_states: Optional[jnp.ndarray] = None,
|
185 |
attention_mask=None,
|
186 |
deterministic: bool = True,
|
187 |
init_cache: bool = False,
|
188 |
output_attentions: bool = False,
|
189 |
):
|
190 |
+
|
191 |
+
# if key_value_states are provided this layer is used as a cross-attention layer
|
192 |
+
# for the decoder
|
193 |
+
is_cross_attention = key_value_states is not None
|
194 |
+
batch_size = hidden_states.shape[0]
|
195 |
+
|
196 |
+
if not is_cross_attention:
|
197 |
+
qkv_out = self.c_attn(hidden_states)
|
198 |
+
query, key, value = jnp.split(qkv_out, 3, axis=2)
|
199 |
+
else:
|
200 |
+
q_out = self.q_attn(hidden_states)
|
201 |
+
(query,) = jnp.split(q_out, 1, axis=2)
|
202 |
+
kv_out = self.c_attn(key_value_states)
|
203 |
+
key, value = jnp.split(kv_out, 2, axis=2)
|
204 |
|
205 |
query = self._split_heads(query)
|
206 |
key = self._split_heads(key)
|
|
|
208 |
|
209 |
query_length, key_length = query.shape[1], key.shape[1]
|
210 |
|
211 |
+
if self.causal:
|
212 |
+
if self.has_variable("cache", "cached_key"):
|
213 |
+
mask_shift = self.variables["cache"]["cache_index"]
|
214 |
+
max_decoder_length = self.variables["cache"]["cached_key"].shape[1]
|
215 |
+
causal_mask = lax.dynamic_slice(
|
216 |
+
self.causal_mask, (0, 0, mask_shift, 0), (1, 1, query_length, max_decoder_length)
|
217 |
+
)
|
218 |
+
else:
|
219 |
+
causal_mask = self.causal_mask[:, :, :query_length, :key_length]
|
220 |
+
causal_mask = jnp.broadcast_to(causal_mask, (batch_size,) + causal_mask.shape[1:])
|
221 |
+
|
222 |
+
# combine masks if needed
|
223 |
+
if attention_mask is not None and self.causal:
|
224 |
+
attention_mask = jnp.broadcast_to(jnp.expand_dims(attention_mask, axis=(-3, -2)), causal_mask.shape)
|
225 |
+
attention_mask = combine_masks(attention_mask, causal_mask)
|
226 |
+
elif self.causal:
|
227 |
+
attention_mask = causal_mask
|
228 |
+
elif attention_mask is not None:
|
229 |
+
attention_mask = jnp.expand_dims(attention_mask, axis=(-3, -2))
|
230 |
|
231 |
dropout_rng = None
|
232 |
if not deterministic and self.config.attn_pdrop > 0.0:
|
|
|
234 |
|
235 |
# During fast autoregressive decoding, we feed one position at a time,
|
236 |
# and cache the keys and values step by step.
|
237 |
+
if self.causal and (self.has_variable("cache", "cached_key") or init_cache):
|
238 |
key, value, attention_mask = self._concatenate_to_cache(key, value, query, attention_mask)
|
239 |
|
240 |
# transform boolean mask into float mask
|
241 |
+
if attention_mask is not None:
|
242 |
+
attention_bias = lax.select(
|
243 |
+
attention_mask > 0,
|
244 |
+
jnp.full(attention_mask.shape, 0.0).astype(self.dtype),
|
245 |
+
jnp.full(attention_mask.shape, -1e4).astype(self.dtype),
|
246 |
+
)
|
247 |
+
else:
|
248 |
+
attention_bias = None
|
249 |
|
250 |
# usual dot product attention
|
251 |
attn_weights = dot_product_attention_weights(
|
|
|
299 |
self.ln_1 = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
300 |
self.attn = FlaxGPT2Attention(self.config, dtype=self.dtype)
|
301 |
self.ln_2 = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
302 |
+
|
303 |
+
if self.config.add_cross_attention:
|
304 |
+
self.crossattention = FlaxGPT2Attention(config=self.config, dtype=self.dtype, causal=False, is_cross_attention=True)
|
305 |
+
self.ln_cross_attn = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
306 |
+
|
307 |
+
project_encoder = getattr(self.config, "project_encoder", None)
|
308 |
+
if project_encoder:
|
309 |
+
self.encoder_projection_ln = nn.LayerNorm(epsilon=self.config.layer_norm_epsilon, dtype=self.dtype)
|
310 |
+
self.encoder_projection_mlp = FlaxGPT2MLP(self.config, self.config.hidden_size, dtype=self.dtype)
|
311 |
+
|
312 |
self.mlp = FlaxGPT2MLP(self.config, inner_dim, dtype=self.dtype)
|
313 |
|
314 |
def __call__(
|
315 |
self,
|
316 |
hidden_states,
|
317 |
attention_mask=None,
|
318 |
+
encoder_hidden_states: Optional[jnp.ndarray] = None,
|
319 |
+
encoder_attention_mask: Optional[jnp.ndarray] = None,
|
320 |
deterministic: bool = True,
|
321 |
init_cache: bool = False,
|
322 |
output_attentions: bool = False,
|
323 |
):
|
324 |
residual = hidden_states
|
325 |
hidden_states = self.ln_1(hidden_states)
|
326 |
+
attn_outputs = self.attn(
|
327 |
hidden_states,
|
328 |
attention_mask=attention_mask,
|
329 |
deterministic=deterministic,
|
|
|
331 |
output_attentions=output_attentions,
|
332 |
)
|
333 |
# residual connection
|
334 |
+
attn_output = attn_outputs[0] # output_attn: a, (attentions)
|
335 |
+
outputs = attn_outputs[1:]
|
336 |
+
# residual connection
|
337 |
hidden_states = attn_output + residual
|
338 |
|
339 |
+
# Cross-Attention Block
|
340 |
+
cross_attn_weights = None
|
341 |
+
if encoder_hidden_states is not None:
|
342 |
+
# add one self-attention block for cross-attention
|
343 |
+
if not hasattr(self, "crossattention"):
|
344 |
+
raise ValueError(
|
345 |
+
f"If `encoder_hidden_states` are passed, {self} has to be instantiated with "
|
346 |
+
"cross-attention layers by setting `config.add_cross_attention=True`"
|
347 |
+
)
|
348 |
+
|
349 |
+
project_encoder = getattr(self.config, "project_encoder", None)
|
350 |
+
if project_encoder:
|
351 |
+
encoder_hidden_states = self.encoder_projection_ln(encoder_hidden_states)
|
352 |
+
feed_forward_hidden_states = self.encoder_projection_mlp(
|
353 |
+
encoder_hidden_states, deterministic=deterministic
|
354 |
+
)
|
355 |
+
# residual connection
|
356 |
+
encoder_hidden_states = feed_forward_hidden_states
|
357 |
+
|
358 |
+
residual = hidden_states
|
359 |
+
hidden_states = self.ln_cross_attn(hidden_states)
|
360 |
+
cross_attn_outputs = self.crossattention(
|
361 |
+
hidden_states,
|
362 |
+
key_value_states=encoder_hidden_states,
|
363 |
+
attention_mask=encoder_attention_mask,
|
364 |
+
deterministic=deterministic,
|
365 |
+
output_attentions=output_attentions,
|
366 |
+
)
|
367 |
+
attn_output = cross_attn_outputs[0]
|
368 |
+
# residual connection
|
369 |
+
hidden_states = residual + attn_output
|
370 |
+
outputs = outputs + cross_attn_outputs[1:] # add cross attentions if we output attention weights
|
371 |
+
|
372 |
residual = hidden_states
|
373 |
hidden_states = self.ln_2(hidden_states)
|
374 |
feed_forward_hidden_states = self.mlp(hidden_states, deterministic=deterministic)
|
375 |
# residual connection
|
376 |
hidden_states = residual + feed_forward_hidden_states
|
377 |
|
378 |
+
outputs = (hidden_states,) + outputs
|
379 |
+
|
380 |
+
return outputs
|
381 |
|
382 |
|
383 |
class FlaxGPT2PreTrainedModel(FlaxPreTrainedModel):
|
|
|
409 |
params_rng, dropout_rng = jax.random.split(rng)
|
410 |
rngs = {"params": params_rng, "dropout": dropout_rng}
|
411 |
|
412 |
+
if self.config.add_cross_attention:
|
413 |
+
encoder_hidden_states = jnp.zeros(input_shape + (self.config.n_embd,))
|
414 |
+
encoder_attention_mask = attention_mask
|
415 |
+
module_init_outputs = self.module.init(
|
416 |
+
rngs, input_ids, attention_mask, position_ids,
|
417 |
+
encoder_hidden_states, encoder_attention_mask, return_dict=False
|
418 |
+
)
|
419 |
+
else:
|
420 |
+
module_init_outputs = self.module.init(
|
421 |
+
rngs, input_ids, attention_mask, position_ids, return_dict=False
|
422 |
+
)
|
423 |
+
|
424 |
+
return module_init_outputs["params"]
|
425 |
|
426 |
def init_cache(self, batch_size, max_length):
|
427 |
r"""
|
|
|
448 |
input_ids,
|
449 |
attention_mask=None,
|
450 |
position_ids=None,
|
451 |
+
encoder_hidden_states: Optional[jnp.ndarray] = None,
|
452 |
+
encoder_attention_mask: Optional[jnp.ndarray] = None,
|
453 |
params: dict = None,
|
454 |
past_key_values: dict = None,
|
455 |
dropout_rng: jax.random.PRNGKey = None,
|
|
|
464 |
)
|
465 |
return_dict = return_dict if return_dict is not None else self.config.return_dict
|
466 |
|
467 |
+
if encoder_hidden_states is not None and encoder_attention_mask is None:
|
468 |
+
batch_size, sequence_length = encoder_hidden_states.shape[:2]
|
469 |
+
encoder_attention_mask = jnp.ones((batch_size, sequence_length))
|
470 |
+
|
471 |
batch_size, sequence_length = input_ids.shape
|
472 |
|
473 |
if position_ids is None:
|
|
|
498 |
jnp.array(input_ids, dtype="i4"),
|
499 |
jnp.array(attention_mask, dtype="i4"),
|
500 |
jnp.array(position_ids, dtype="i4"),
|
501 |
+
encoder_hidden_states,
|
502 |
+
encoder_attention_mask,
|
503 |
not train,
|
504 |
False,
|
505 |
output_attentions,
|
|
|
534 |
self,
|
535 |
hidden_states,
|
536 |
attention_mask=None,
|
537 |
+
encoder_hidden_states: Optional[jnp.ndarray] = None,
|
538 |
+
encoder_attention_mask: Optional[jnp.ndarray] = None,
|
539 |
deterministic: bool = True,
|
540 |
init_cache: bool = False,
|
541 |
output_attentions: bool = False,
|
|
|
544 |
):
|
545 |
all_attentions = () if output_attentions else None
|
546 |
all_hidden_states = () if output_hidden_states else None
|
547 |
+
all_cross_attentions = () if (output_attentions and encoder_hidden_states is not None) else None
|
548 |
|
549 |
for block in self.blocks:
|
550 |
if output_hidden_states:
|
|
|
553 |
layer_outputs = block(
|
554 |
hidden_states,
|
555 |
attention_mask,
|
556 |
+
encoder_hidden_states=encoder_hidden_states,
|
557 |
+
encoder_attention_mask=encoder_attention_mask,
|
558 |
deterministic=deterministic,
|
559 |
init_cache=init_cache,
|
560 |
output_attentions=output_attentions,
|
|
|
564 |
if output_attentions:
|
565 |
all_attentions += (layer_outputs[1],)
|
566 |
|
567 |
+
if encoder_hidden_states is not None:
|
568 |
+
all_cross_attentions += (layer_outputs[2],)
|
569 |
+
|
570 |
if output_hidden_states:
|
571 |
all_hidden_states += (hidden_states,)
|
572 |
|
573 |
+
outputs = [hidden_states, all_hidden_states, all_attentions, all_cross_attentions]
|
574 |
|
575 |
if not return_dict:
|
576 |
return tuple(v for v in outputs if v is not None)
|
577 |
|
578 |
+
return FlaxBaseModelOutputWithPastAndCrossAttentions(
|
579 |
last_hidden_state=hidden_states,
|
|
|
580 |
hidden_states=all_hidden_states,
|
581 |
attentions=all_attentions,
|
582 |
+
cross_attentions=all_cross_attentions,
|
583 |
)
|
584 |
|
585 |
|
|
|
611 |
input_ids,
|
612 |
attention_mask,
|
613 |
position_ids,
|
614 |
+
encoder_hidden_states: Optional[jnp.ndarray] = None,
|
615 |
+
encoder_attention_mask: Optional[jnp.ndarray] = None,
|
616 |
deterministic=True,
|
617 |
init_cache: bool = False,
|
618 |
output_attentions: bool = False,
|
|
|
628 |
outputs = self.h(
|
629 |
hidden_states,
|
630 |
attention_mask,
|
631 |
+
encoder_hidden_states,
|
632 |
+
encoder_attention_mask,
|
633 |
deterministic=deterministic,
|
634 |
init_cache=init_cache,
|
635 |
output_attentions=output_attentions,
|
|
|
643 |
if not return_dict:
|
644 |
return (hidden_states,) + outputs[1:]
|
645 |
|
646 |
+
return FlaxBaseModelOutputWithPastAndCrossAttentions(
|
647 |
last_hidden_state=hidden_states,
|
648 |
hidden_states=outputs.hidden_states,
|
649 |
attentions=outputs.attentions,
|
650 |
+
cross_attentions=outputs.cross_attentions,
|
651 |
)
|
652 |
|
653 |
|
|
|
660 |
|
661 |
|
662 |
append_call_sample_docstring(
|
663 |
+
FlaxGPT2Model, _TOKENIZER_FOR_DOC, _CHECKPOINT_FOR_DOC, FlaxBaseModelOutputWithPastAndCrossAttentions, _CONFIG_FOR_DOC
|
664 |
)
|
665 |
|
666 |
|
|
|
682 |
input_ids,
|
683 |
attention_mask,
|
684 |
position_ids,
|
685 |
+
encoder_hidden_states: Optional[jnp.ndarray] = None,
|
686 |
+
encoder_attention_mask: Optional[jnp.ndarray] = None,
|
687 |
deterministic: bool = True,
|
688 |
init_cache: bool = False,
|
689 |
output_attentions: bool = False,
|
|
|
694 |
input_ids,
|
695 |
attention_mask,
|
696 |
position_ids,
|
697 |
+
encoder_hidden_states,
|
698 |
+
encoder_attention_mask,
|
699 |
deterministic=deterministic,
|
700 |
init_cache=init_cache,
|
701 |
output_attentions=output_attentions,
|
|
|
714 |
if not return_dict:
|
715 |
return (lm_logits,) + outputs[1:]
|
716 |
|
717 |
+
return FlaxCausalLMOutputWithCrossAttentions(
|
718 |
+
logits=lm_logits,
|
719 |
+
hidden_states=outputs.hidden_states,
|
720 |
+
attentions=outputs.attentions,
|
721 |
+
cross_attentions=outputs.cross_attentions
|
722 |
+
)
|
723 |
|
724 |
@add_start_docstrings(
|
725 |
"""
|
|
|
759 |
|
760 |
|
761 |
append_call_sample_docstring(
|
762 |
+
FlaxGPT2LMHeadModel, _TOKENIZER_FOR_DOC, _CHECKPOINT_FOR_DOC, FlaxCausalLMOutputWithCrossAttentions, _CONFIG_FOR_DOC
|
763 |
)
|