modeling_tf_esm.py

# coding=utf-8
# Copyright 2022 Meta and The HuggingFace Inc. team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
""" PyTorch ESM model."""


from __future__ import annotations

import os
from typing import Optional, Tuple, Union

import numpy as np
import tensorflow as tf
from tensorflow.keras.activations import gelu
from tensorflow.keras.layers import Dense, Dropout, Embedding, Layer, LayerNormalization

from ...file_utils import add_code_sample_docstrings, add_start_docstrings, add_start_docstrings_to_model_forward
from ...modeling_tf_outputs import (
    TFBaseModelOutputWithPastAndCrossAttentions,
    TFBaseModelOutputWithPoolingAndCrossAttentions,
    TFMaskedLMOutput,
    TFSequenceClassifierOutput,
    TFTokenClassifierOutput,
)
from ...modeling_tf_utils import (
    TFMaskedLanguageModelingLoss,
    TFModelInputType,
    TFPreTrainedModel,
    TFSequenceClassificationLoss,
    TFTokenClassificationLoss,
    get_initializer,
    shape_list,
    unpack_inputs,
)
from ...tf_utils import check_embeddings_within_bounds, stable_softmax
from ...utils import logging
from .configuration_esm import EsmConfig


logger = logging.get_logger(__name__)

_CHECKPOINT_FOR_DOC = "facebook/esm2_t6_8M_UR50D"
_CONFIG_FOR_DOC = "EsmConfig"

TF_ESM_PRETRAINED_MODEL_ARCHIVE_LIST = [
    "facebook/esm2_t6_8M_UR50D",
    "facebook/esm2_t12_35M_UR50D",
    # This is not a complete list of all ESM models!
    # See all ESM models at https://huggingface.co/models?filter=esm
]


def rotate_half(x):
    x1, x2 = tf.split(x, 2, axis=-1)
    return tf.concat((-x2, x1), axis=-1)


def apply_rotary_pos_emb(x, cos, sin):
    cos = cos[:, :, : tf.shape(x)[-2], :]
    sin = sin[:, :, : tf.shape(x)[-2], :]

    return (x * cos) + (rotate_half(x) * sin)


def symmetrize(x):
    "Make layer symmetric in final two dimensions, used for contact prediction."
    return x + tf.linalg.matrix_transpose(x)  # Transposes last two dimensions only


def average_product_correct(x):
    "Perform average product correct, used for contact prediction."
    a1 = tf.reduce_sum(x, -1, keepdims=True)
    a2 = tf.reduce_sum(x, -2, keepdims=True)
    a12 = tf.reduce_sum(x, (-1, -2), keepdims=True)

    avg = a1 * a2
    avg = avg / a12
    normalized = x - avg
    return normalized


class TFRotaryEmbedding(Layer):
    """
    Rotary position embeddings based on those in
    [RoFormer](https://huggingface.co/docs/transformers/model_doc/roformer). Query and keys are transformed by rotation
    matrices which depend on their relative positions.
    """

    def __init__(self, dim: int, name=None):
        super().__init__(name=name)
        # Matt: The PyTorch version of this layer does a lot of work to cache values, but we just rely on TF compilation
        # and/or XLA to sort out constants like that. It actually may not seem like this layer needs to be stateful at
        # all when we benefit from TF compilation, but it does. The reason is that self.inv_freq is a buffer in the
        # original implementation, but all the shared ESM checkpoints were trained with fp16 params. This means that
        # the inv_freq tensor was stored as a float16, and we need to replicate those lower-precision values or our
        # models give different outputs from the original.
        self.dim = dim

    def build(self, input_shape):
        super().build(input_shape)
        self.inv_freq = self.add_weight(
            "inv_freq", shape=(self.dim // 2,), dtype=tf.float32, initializer=get_initializer(1.0), trainable=False
        )
        self.inv_freq.assign(
            1.0 / (10000 ** (tf.range(start=0, limit=self.dim, delta=2, dtype=tf.float32) / self.dim))
        )

    def _compute_cos_sin(self, x, seq_dimension=2):
        seq_len = tf.shape(x)[seq_dimension]

        t = tf.range(seq_len, dtype=self.inv_freq.dtype)
        freqs = tf.einsum("i, j -> ij", t, self.inv_freq)  # Outer multiplication
        emb = tf.concat((freqs, freqs), axis=-1)[None, None, :, :]

        return tf.cos(emb), tf.sin(emb)

    def call(self, q: tf.Tensor, k: tf.Tensor) -> Tuple[tf.Tensor, tf.Tensor]:
        cos_emb, sin_emb = self._compute_cos_sin(k, seq_dimension=-2)

        return (
            apply_rotary_pos_emb(q, cos_emb, sin_emb),
            apply_rotary_pos_emb(k, cos_emb, sin_emb),
        )


class TFEsmContactPredictionHead(Layer):
    """Performs symmetrization, apc, and computes a logistic regression on the output features"""

    def __init__(
        self,
        in_features: int,
        bias=True,
        eos_idx: int = 2,
        name=None,
    ):
        super().__init__(name=name)
        self.eos_idx = eos_idx
        self.in_features = in_features
        self.regression = Dense(1, use_bias=bias, activation="sigmoid", name="regression")

    def build(self, input_shape):
        super().build(input_shape)
        with tf.name_scope("regression"):
            self.regression.build((None, self.in_features))

    def call(self, tokens, attentions):
        # remove eos token attentions
        eos_mask = tf.cast(tokens != self.eos_idx, attentions.dtype)
        eos_mask = tf.expand_dims(eos_mask, 1) * tf.expand_dims(eos_mask, 2)
        attentions = attentions * eos_mask[:, None, None, :, :]
        attentions = attentions[..., :-1, :-1]
        # remove cls token attentions
        attentions = attentions[..., 1:, 1:]
        batch_size, layers, heads, seqlen, _ = shape_list(attentions)
        attentions = tf.reshape(attentions, (batch_size, layers * heads, seqlen, seqlen))

        # features: batch x channels x tokens x tokens (symmetric)
        attentions = average_product_correct(symmetrize(attentions))
        attentions = tf.transpose(attentions, perm=(0, 2, 3, 1))
        return tf.squeeze(self.regression(attentions), 3)


class TFEsmEmbeddings(Layer):
    """
    Same as BertEmbeddings with a tiny tweak for positional embeddings indexing.
    """

    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.word_embeddings = Embedding(
            config.vocab_size,
            config.hidden_size,
            embeddings_initializer=get_initializer(config.initializer_range),
            name="word_embeddings",
        )
        self.position_embeddings = Embedding(
            config.max_position_embeddings,
            config.hidden_size,
            embeddings_initializer=get_initializer(config.initializer_range),
            name="position_embeddings",
        )

        if config.emb_layer_norm_before:
            self.layer_norm = LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm")
        else:
            self.layer_norm = None
        # Matt: I think this line was copied incorrectly from BERT, disabling for now
        # self.dropout = Dropout(config.hidden_dropout_prob)
        self.position_embedding_type = getattr(config, "position_embedding_type", "absolute")

        self.position_ids = tf.range(config.max_position_embeddings)[None, :]

        self.padding_idx = config.pad_token_id
        self.token_dropout = config.token_dropout
        self.mask_token_id = config.mask_token_id
        self.config = config

    def call(
        self, input_ids=None, attention_mask=None, position_ids=None, inputs_embeds=None, past_key_values_length=0
    ):
        if position_ids is None:
            if input_ids is not None:
                # Create the position ids from the input token ids. Any padded tokens remain padded.
                position_ids = create_position_ids_from_input_ids(input_ids, self.padding_idx, past_key_values_length)
            else:
                position_ids = self.create_position_ids_from_inputs_embeds(inputs_embeds)

        if inputs_embeds is None:
            check_embeddings_within_bounds(input_ids, self.config.vocab_size)
            inputs_embeds = self.word_embeddings(input_ids)

        # Note that if we want to support ESM-1 (not 1b!) in future then we need to support an
        # embedding_scale factor here.
        embeddings = inputs_embeds

        # Matt: ESM has the option to handle masking in MLM in a slightly unusual way. If the token_dropout
        # flag is False then it is handled in the same was as BERT/RoBERTa. If it is set to True, however,
        # masked tokens are treated as if they were selected for input dropout and zeroed out.
        # This "mask-dropout" is compensated for when masked tokens are not present, by scaling embeddings by
        # a factor of (fraction of unmasked tokens during training) / (fraction of unmasked tokens in sample).
        # This is analogous to the way that dropout layers scale down outputs during evaluation when not
        # actually dropping out values (or, equivalently, scale up their un-dropped outputs in training).
        if self.token_dropout:
            embeddings = tf.where((input_ids == self.mask_token_id)[:, :, None], 0.0, embeddings)
            mask_ratio_train = 0.15 * 0.8  # Hardcoded as the ratio used in all ESM model training runs
            src_lengths = tf.cast(tf.reduce_sum(attention_mask, axis=-1), tf.float32)
            masked_tokens = input_ids == self.mask_token_id
            mask_ratio_observed = tf.math.count_nonzero(masked_tokens, dtype=tf.float32, axis=-1) / src_lengths
            embeddings = embeddings * (1 - mask_ratio_train) / (1 - mask_ratio_observed)[:, None, None]

        if self.position_embedding_type == "absolute":
            position_embeddings = self.position_embeddings(position_ids)
            embeddings += position_embeddings

        if self.layer_norm is not None:
            embeddings = self.layer_norm(embeddings)
        if attention_mask is not None:
            embeddings = embeddings * tf.cast(tf.expand_dims(attention_mask, -1), embeddings.dtype)
        # Matt: I think this line was copied incorrectly from BERT, disabling it for now.
        # embeddings = self.dropout(embeddings)
        return embeddings

    def create_position_ids_from_inputs_embeds(self, inputs_embeds):
        """
        We are provided embeddings directly. We cannot infer which are padded so just generate sequential position ids.

        Args:
            inputs_embeds: tf.Tensor

        Returns: tf.Tensor
        """
        input_shape = shape_list(inputs_embeds)[:-1]
        sequence_length = input_shape[1]

        position_ids = tf.range(
            start=self.padding_idx + 1, limit=sequence_length + self.padding_idx + 1, dtype=tf.int64
        )
        return tf.broadcast_to(tf.expand_dims(position_ids, 0), input_shape)


class TFEsmSelfAttention(Layer):
    def __init__(self, config, position_embedding_type=None, name=None):
        super().__init__(name=name)
        if config.hidden_size % config.num_attention_heads != 0 and not hasattr(config, "embedding_size"):
            raise ValueError(
                f"The hidden size ({config.hidden_size}) is not a multiple of the number of attention "
                f"heads ({config.num_attention_heads})"
            )

        self.num_attention_heads = config.num_attention_heads
        self.attention_head_size = int(config.hidden_size / config.num_attention_heads)
        self.all_head_size = self.num_attention_heads * self.attention_head_size

        self.query = Dense(
            self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="query"
        )
        self.key = Dense(self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="key")
        self.value = Dense(
            self.all_head_size, kernel_initializer=get_initializer(config.initializer_range), name="value"
        )

        self.dropout = Dropout(config.attention_probs_dropout_prob)
        self.position_embedding_type = position_embedding_type or getattr(
            config, "position_embedding_type", "absolute"
        )
        self.rotary_embeddings = None
        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
            self.max_position_embeddings = config.max_position_embeddings
            self.distance_embedding = Embedding(
                2 * config.max_position_embeddings - 1,
                self.attention_head_size,
                embeddings_initializer=get_initializer(config.initializer_range),
            )
        elif self.position_embedding_type == "rotary":
            self.rotary_embeddings = TFRotaryEmbedding(dim=self.attention_head_size, name="rotary_embeddings")

        self.is_decoder = config.is_decoder

    def transpose_for_scores(self, x: tf.Tensor) -> tf.Tensor:
        new_x_shape = shape_list(x)[:-1] + [self.num_attention_heads, self.attention_head_size]
        x = tf.reshape(x, new_x_shape)
        return tf.transpose(x, perm=(0, 2, 1, 3))

    def call(
        self,
        hidden_states: tf.Tensor,
        attention_mask: tf.Tensor | None = None,
        head_mask: tf.Tensor | None = None,
        encoder_hidden_states: tf.Tensor | None = None,
        encoder_attention_mask: tf.Tensor | None = None,
        past_key_value: Tuple[Tuple[tf.Tensor]] | None = None,
        output_attentions: Optional[bool] = False,
        training: bool = False,
    ) -> Tuple[tf.Tensor]:
        mixed_query_layer = self.query(hidden_states)

        # If this is instantiated as a cross-attention module, the keys
        # and values come from an encoder; the attention mask needs to be
        # such that the encoder's padding tokens are not attended to.
        is_cross_attention = encoder_hidden_states is not None

        if is_cross_attention and past_key_value is not None:
            # reuse k,v, cross_attentions
            key_layer = past_key_value[0]
            value_layer = past_key_value[1]
            attention_mask = encoder_attention_mask
        elif is_cross_attention:
            key_layer = self.transpose_for_scores(self.key(encoder_hidden_states))
            value_layer = self.transpose_for_scores(self.value(encoder_hidden_states))
            attention_mask = encoder_attention_mask
        elif past_key_value is not None:
            key_layer = self.transpose_for_scores(self.key(hidden_states))
            value_layer = self.transpose_for_scores(self.value(hidden_states))
            key_layer = tf.concat([past_key_value[0], key_layer], axis=2)
            value_layer = tf.concat([past_key_value[1], value_layer], axis=2)
        else:
            key_layer = self.transpose_for_scores(self.key(hidden_states))
            value_layer = self.transpose_for_scores(self.value(hidden_states))

        query_layer = self.transpose_for_scores(mixed_query_layer)

        # Matt: Our BERT model (which this code was derived from) scales attention logits down by sqrt(head_dim).
        # ESM scales the query down by the same factor instead. Modulo numerical stability these are equivalent,
        # but not when rotary embeddings get involved. Therefore, we scale the query here to match the original
        # ESM code and fix rotary embeddings.
        query_layer = query_layer * self.attention_head_size**-0.5

        if self.is_decoder:
            # if cross_attention save Tuple(tf.Tensor, tf.Tensor) of all cross attention key/value_states.
            # Further calls to cross_attention layer can then reuse all cross-attention
            # key/value_states (first "if" case)
            # if uni-directional self-attention (decoder) save Tuple(tf.Tensor, tf.Tensor) of
            # all previous decoder key/value_states. Further calls to uni-directional self-attention
            # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case)
            # if encoder bi-directional self-attention `past_key_value` is always `None`
            past_key_value = (key_layer, value_layer)

        if self.position_embedding_type == "rotary":
            query_layer, key_layer = self.rotary_embeddings(query_layer, key_layer)

        # Take the dot product between "query" and "key" to get the raw attention scores.
        attention_scores = tf.matmul(query_layer, key_layer, transpose_b=True)

        if self.position_embedding_type == "relative_key" or self.position_embedding_type == "relative_key_query":
            seq_length = shape_list(hidden_states)[1]
            position_ids_l = tf.expand_dims(tf.range(seq_length, dtype=tf.int64), -1)
            position_ids_r = tf.expand_dims(tf.range(seq_length, dtype=tf.int64), 0)
            distance = position_ids_l - position_ids_r
            positional_embedding = self.distance_embedding(distance + self.max_position_embeddings - 1)
            positional_embedding = tf.cast(positional_embedding, query_layer.dtype)  # fp16 compatibility

            if self.position_embedding_type == "relative_key":
                relative_position_scores = tf.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
                attention_scores = attention_scores + relative_position_scores
            elif self.position_embedding_type == "relative_key_query":
                relative_position_scores_query = tf.einsum("bhld,lrd->bhlr", query_layer, positional_embedding)
                relative_position_scores_key = tf.einsum("bhrd,lrd->bhlr", key_layer, positional_embedding)
                attention_scores = attention_scores + relative_position_scores_query + relative_position_scores_key

        if attention_mask is not None:
            # Apply the attention mask is (precomputed for all layers in EsmModel forward() function)
            attention_scores = attention_scores + attention_mask

        # Normalize the attention scores to probabilities.
        attention_probs = stable_softmax(attention_scores, axis=-1)

        # This is actually dropping out entire tokens to attend to, which might
        # seem a bit unusual, but is taken from the original Transformer paper.
        attention_probs = self.dropout(attention_probs, training=training)

        # Mask heads if we want to
        if head_mask is not None:
            attention_probs = attention_probs * head_mask

        context_layer = attention_probs @ value_layer

        context_layer = tf.transpose(context_layer, perm=(0, 2, 1, 3))
        new_context_layer_shape = shape_list(context_layer)[:-2] + [self.all_head_size]
        context_layer = tf.reshape(context_layer, new_context_layer_shape)

        outputs = (context_layer, attention_probs) if output_attentions else (context_layer,)

        if self.is_decoder:
            outputs = outputs + (past_key_value,)
        return outputs


class TFEsmSelfOutput(Layer):
    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.dense = Dense(
            config.hidden_size, kernel_initializer=get_initializer(config.initializer_range), name="dense"
        )
        self.dropout = Dropout(config.hidden_dropout_prob)

    def call(self, hidden_states, input_tensor, training=False):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.dropout(hidden_states, training=training)
        hidden_states += input_tensor
        return hidden_states


class TFEsmAttention(Layer):
    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.self = TFEsmSelfAttention(config, name="self")
        self.output_layer = TFEsmSelfOutput(config, name="output")
        self.pruned_heads = set()
        self.LayerNorm = LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm")

    def prune_heads(self, heads):
        raise NotImplementedError

    def call(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_value=None,
        output_attentions=False,
        training=False,
    ):
        hidden_states_ln = self.LayerNorm(hidden_states)
        self_outputs = self.self(
            hidden_states_ln,
            attention_mask,
            head_mask,
            encoder_hidden_states,
            encoder_attention_mask,
            past_key_value,
            output_attentions,
            training,
        )
        attention_output = self.output_layer(self_outputs[0], hidden_states)
        outputs = (attention_output,) + self_outputs[1:]  # add attentions if we output them
        return outputs


class TFEsmIntermediate(tf.keras.layers.Layer):
    def __init__(self, config: EsmConfig, **kwargs):
        super().__init__(**kwargs)

        self.dense = tf.keras.layers.Dense(
            units=config.intermediate_size,
            kernel_initializer=get_initializer(config.initializer_range),
            name="dense",
        )

    def call(self, hidden_states: tf.Tensor) -> tf.Tensor:
        hidden_states = self.dense(inputs=hidden_states)
        hidden_states = tf.nn.gelu(hidden_states)
        return hidden_states


class TFEsmOutput(Layer):
    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.dense = Dense(
            config.hidden_size, kernel_initializer=get_initializer(config.initializer_range), name="dense"
        )
        self.dropout = Dropout(config.hidden_dropout_prob)

    def call(self, hidden_states, input_tensor, training=False):
        hidden_states = self.dense(hidden_states)
        hidden_states = self.dropout(hidden_states, training=training)
        hidden_states += input_tensor
        return hidden_states


class TFEsmLayer(Layer):
    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.chunk_size_feed_forward = config.chunk_size_feed_forward
        self.seq_len_dim = 1
        self.attention = TFEsmAttention(config, name="attention")
        self.is_decoder = config.is_decoder
        self.add_cross_attention = config.add_cross_attention
        if self.add_cross_attention:
            if not self.is_decoder:
                raise RuntimeError(f"{self} should be used as a decoder model if cross attention is added")
            self.crossattention = TFEsmAttention(config)
        self.intermediate = TFEsmIntermediate(config, name="intermediate")
        self.output_layer = TFEsmOutput(config, name="output")
        self.LayerNorm = LayerNormalization(epsilon=config.layer_norm_eps, name="LayerNorm")

    def call(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_value=None,
        output_attentions=False,
        training=False,
    ):
        # decoder uni-directional self-attention cached key/values tuple is at positions 1,2
        self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None
        self_attention_outputs = self.attention(
            hidden_states,
            attention_mask,
            head_mask,
            output_attentions=output_attentions,
            past_key_value=self_attn_past_key_value,
            training=training,
        )
        attention_output = self_attention_outputs[0]

        # if decoder, the last output is tuple of self-attn cache
        if self.is_decoder:
            outputs = self_attention_outputs[1:-1]
            present_key_value = self_attention_outputs[-1]
        else:
            outputs = self_attention_outputs[1:]  # add self attentions if we output attention weights

        cross_attn_present_key_value = None
        if self.is_decoder and encoder_hidden_states is not None:
            if not hasattr(self, "crossattention"):
                raise AttributeError(
                    f"If `encoder_hidden_states` are passed, {self} has to be instantiated"
                    " with cross-attention layers by setting `config.add_cross_attention=True`"
                )

            # cross_attn cached key/values tuple is at positions 3,4 of past_key_value tuple
            cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None
            cross_attention_outputs = self.crossattention(
                attention_output,
                attention_mask,
                head_mask,
                encoder_hidden_states,
                encoder_attention_mask,
                cross_attn_past_key_value,
                output_attentions,
                training=training,
            )
            attention_output = cross_attention_outputs[0]
            outputs = outputs + cross_attention_outputs[1:-1]  # add cross attentions if we output attention weights

            # add cross-attn cache to positions 3,4 of present_key_value tuple
            cross_attn_present_key_value = cross_attention_outputs[-1]
            present_key_value = present_key_value + cross_attn_present_key_value

        layernorm_output = self.LayerNorm(attention_output)
        intermediate_output = self.intermediate(hidden_states=layernorm_output)
        layer_output = self.output_layer(
            hidden_states=intermediate_output, input_tensor=attention_output, training=training
        )
        outputs = (layer_output,) + outputs  # add attentions if we output them

        # if decoder, return the attn key/values as the last output
        if self.is_decoder:
            outputs = outputs + (present_key_value,)

        return outputs


class TFEsmEncoder(Layer):
    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.config = config
        self.layer = [TFEsmLayer(config, name=f"layer_._{i}") for i in range(config.num_hidden_layers)]
        self.emb_layer_norm_after = LayerNormalization(epsilon=config.layer_norm_eps, name="emb_layer_norm_after")

    def call(
        self,
        hidden_states,
        attention_mask=None,
        head_mask=None,
        encoder_hidden_states=None,
        encoder_attention_mask=None,
        past_key_values=None,
        use_cache=None,
        output_attentions=False,
        output_hidden_states=False,
        return_dict=True,
        training=False,
    ):
        all_hidden_states = () if output_hidden_states else None
        all_self_attentions = () if output_attentions else None
        all_cross_attentions = () if output_attentions and self.config.add_cross_attention else None

        next_decoder_cache = () if use_cache else None
        for i, layer_module in enumerate(self.layer):
            if output_hidden_states:
                all_hidden_states = all_hidden_states + (hidden_states,)

            layer_head_mask = head_mask[i] if head_mask is not None else None
            past_key_value = past_key_values[i] if past_key_values is not None else None

            layer_outputs = layer_module(
                hidden_states,
                attention_mask,
                layer_head_mask,
                encoder_hidden_states,
                encoder_attention_mask,
                past_key_value,
                output_attentions,
                training,
            )

            hidden_states = layer_outputs[0]
            if use_cache:
                next_decoder_cache += (layer_outputs[-1],)
            if output_attentions:
                all_self_attentions = all_self_attentions + (layer_outputs[1],)
                if self.config.add_cross_attention:
                    all_cross_attentions = all_cross_attentions + (layer_outputs[2],)

        if self.emb_layer_norm_after:
            hidden_states = self.emb_layer_norm_after(hidden_states)

        if output_hidden_states:
            all_hidden_states = all_hidden_states + (hidden_states,)

        if not return_dict:
            return tuple(
                v
                for v in [
                    hidden_states,
                    next_decoder_cache,
                    all_hidden_states,
                    all_self_attentions,
                    all_cross_attentions,
                ]
                if v is not None
            )
        return TFBaseModelOutputWithPastAndCrossAttentions(
            last_hidden_state=hidden_states,
            past_key_values=next_decoder_cache,
            hidden_states=all_hidden_states,
            attentions=all_self_attentions,
            cross_attentions=all_cross_attentions,
        )


# Copied from transformers.models.bert.modeling_tf_bert.TFBertPooler with Bert->Esm
class TFEsmPooler(tf.keras.layers.Layer):
    def __init__(self, config: EsmConfig, **kwargs):
        super().__init__(**kwargs)

        self.dense = tf.keras.layers.Dense(
            units=config.hidden_size,
            kernel_initializer=get_initializer(config.initializer_range),
            activation="tanh",
            name="dense",
        )

    def call(self, hidden_states: tf.Tensor) -> tf.Tensor:
        # We "pool" the model by simply taking the hidden state corresponding
        # to the first token.
        first_token_tensor = hidden_states[:, 0]
        pooled_output = self.dense(inputs=first_token_tensor)

        return pooled_output


class TFEsmPreTrainedModel(TFPreTrainedModel):
    """
    An abstract class to handle weights initialization and a simple interface for downloading and loading pretrained
    models.
    """

    config_class = EsmConfig
    base_model_prefix = "esm"


ESM_START_DOCSTRING = r"""

    This model inherits from [`TFPreTrainedModel`]. Check the superclass documentation for the generic methods the
    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
    etc.)

    This model is also a Keras [Model](https://www.tensorflow.org/api_docs/python/tf/keras/Model) subclass. Use it as a
    regular Keras model and refer to the TF/Keras documentation for all matters related to general usage and behavior.

    Parameters:
        config ([`EsmConfig`]): Model configuration class with all the parameters of the
            model. Initializing with a config file does not load the weights associated with the model, only the
            configuration. Check out the [`~TFPreTrainedModel.from_pretrained`] method to load the model weights.
"""

ESM_INPUTS_DOCSTRING = r"""
    Args:
        input_ids (`tf.Tensor` of shape `({0})`):
            Indices of input sequence tokens in the vocabulary.

            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)
        attention_mask (`tf.Tensor` of shape `({0})`, *optional*):
            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:

            - 1 for tokens that are **not masked**,
            - 0 for tokens that are **masked**.

            [What are attention masks?](../glossary#attention-mask)
        position_ids (`tf.Tensor` of shape `({0})`, *optional*):
            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
            config.max_position_embeddings - 1]`.

            [What are position IDs?](../glossary#position-ids)
        head_mask (`tf.Tensor` of shape `(num_heads,)` or `(num_layers, num_heads)`, *optional*):
            Mask to nullify selected heads of the self-attention modules. Mask values selected in `[0, 1]`:

            - 1 indicates the head is **not masked**,
            - 0 indicates the head is **masked**.

        inputs_embeds (`tf.Tensor` of shape `({0}, hidden_size)`, *optional*):
            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
            model's internal embedding lookup matrix.
        output_attentions (`bool`, *optional*):
            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
            tensors for more detail.
        output_hidden_states (`bool`, *optional*):
            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
            more detail.
        return_dict (`bool`, *optional*):
            Whether or not to return a [`~file_utils.ModelOutput`] instead of a plain tuple.
"""


@add_start_docstrings(
    "The bare ESM Model transformer outputting raw hidden-states without any specific head on top.",
    ESM_START_DOCSTRING,
)
class TFEsmMainLayer(Layer):
    """

    The model can behave as an encoder (with only self-attention) as well as a decoder, in which case a layer of
    cross-attention is added between the self-attention layers, following the architecture described in [Attention is
    all you need](https://arxiv.org/abs/1706.03762) by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit,
    Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin.

    To behave as an decoder the model needs to be initialized with the `is_decoder` argument of the configuration set
    to `True`. To be used in a Seq2Seq model, the model needs to initialized with both `is_decoder` argument and
    `add_cross_attention` set to `True`; an `encoder_hidden_states` is then expected as an input to the forward pass.
    """

    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config, add_pooling_layer=True, name=None, **kwargs):
        super().__init__(name=name, **kwargs)

        self.config = config
        self.is_decoder = config.is_decoder

        self.embeddings = TFEsmEmbeddings(config, name="embeddings")
        self.encoder = TFEsmEncoder(config, name="encoder")
        self.pooler = TFEsmPooler(config, name="pooler") if add_pooling_layer else None

        self.contact_head = TFEsmContactPredictionHead(
            in_features=self.config.num_hidden_layers * self.config.num_attention_heads, bias=True, name="contact_head"
        )

    def build(self, input_shape):
        super().build(input_shape)
        with tf.name_scope("contact_head"):
            self.contact_head.build(input_shape)

    def get_input_embeddings(self):
        return self.embeddings.word_embeddings

    def set_input_embeddings(self, value: tf.Variable):
        self.embeddings.word_embeddings.weight = value
        self.embeddings.vocab_size = shape_list(value)[0]

    def _prune_heads(self, heads_to_prune):
        raise NotImplementedError

    def call(
        self,
        input_ids: TFModelInputType | None = None,
        attention_mask: np.ndarray | tf.Tensor | None = None,
        position_ids: np.ndarray | tf.Tensor | None = None,
        head_mask: np.ndarray | tf.Tensor | None = None,
        inputs_embeds: np.ndarray | tf.Tensor | None = None,
        encoder_hidden_states: np.ndarray | tf.Tensor | None = None,
        encoder_attention_mask: np.ndarray | tf.Tensor | None = None,
        past_key_values: Optional[Tuple[Tuple[Union[np.ndarray, tf.Tensor]]]] = None,
        use_cache: Optional[bool] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        training: bool = False,
    ) -> Union[TFBaseModelOutputWithPoolingAndCrossAttentions, Tuple[tf.Tensor]]:
        if not self.config.is_decoder:
            use_cache = False

        if input_ids is not None and inputs_embeds is not None:
            raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time")
        elif input_ids is not None:
            input_shape = shape_list(input_ids)
        elif inputs_embeds is not None:
            input_shape = shape_list(inputs_embeds)[:-1]
        else:
            raise ValueError("You have to specify either input_ids or inputs_embeds")

        batch_size, seq_length = input_shape

        if past_key_values is None:
            past_key_values_length = 0
            past_key_values = [None] * len(self.encoder.layer)
        else:
            past_key_values_length = shape_list(past_key_values[0][0])[-2]

        if attention_mask is None:
            attention_mask = tf.fill(dims=(batch_size, seq_length + past_key_values_length), value=1)

        embedding_output = self.embeddings(
            input_ids=input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            inputs_embeds=inputs_embeds,
            past_key_values_length=past_key_values_length,
            training=training,
        )

        # We create a 3D attention mask from a 2D tensor mask.
        # Sizes are [batch_size, 1, 1, to_seq_length]
        # So we can broadcast to [batch_size, num_heads, from_seq_length, to_seq_length]
        # this attention mask is more simple than the triangular masking of causal attention
        # used in OpenAI GPT, we just need to prepare the broadcast dimension here.
        attention_mask_shape = shape_list(attention_mask)

        mask_seq_length = seq_length + past_key_values_length
        # Copied from `modeling_tf_t5.py`
        # Provided a padding mask of dimensions [batch_size, mask_seq_length]
        # - if the model is a decoder, apply a causal mask in addition to the padding mask
        # - if the model is an encoder, make the mask broadcastable to [batch_size, num_heads, mask_seq_length, mask_seq_length]
        if self.is_decoder:
            seq_ids = tf.range(mask_seq_length)
            causal_mask = tf.less_equal(
                tf.tile(seq_ids[None, None, :], (batch_size, mask_seq_length, 1)),
                seq_ids[None, :, None],
            )
            causal_mask = tf.cast(causal_mask, dtype=attention_mask.dtype)
            extended_attention_mask = causal_mask * attention_mask[:, None, :]
            attention_mask_shape = shape_list(extended_attention_mask)
            extended_attention_mask = tf.reshape(
                extended_attention_mask, (attention_mask_shape[0], 1, attention_mask_shape[1], attention_mask_shape[2])
            )
            if past_key_values[0] is not None:
                # attention_mask needs to be sliced to the shape `[batch_size, 1, from_seq_length - cached_seq_length, to_seq_length]
                extended_attention_mask = extended_attention_mask[:, :, -seq_length:, :]
        else:
            extended_attention_mask = tf.reshape(
                attention_mask, (attention_mask_shape[0], 1, 1, attention_mask_shape[1])
            )

        # Since attention_mask is 1.0 for positions we want to attend and 0.0 for
        # masked positions, this operation will create a tensor which is 0.0 for
        # positions we want to attend and -10000.0 for masked positions.
        # Since we are adding it to the raw scores before the softmax, this is
        # effectively the same as removing these entirely.
        extended_attention_mask = tf.cast(extended_attention_mask, dtype=embedding_output.dtype)
        one_cst = tf.constant(1.0, dtype=embedding_output.dtype)
        ten_thousand_cst = tf.constant(-10000.0, dtype=embedding_output.dtype)
        extended_attention_mask = tf.multiply(tf.subtract(one_cst, extended_attention_mask), ten_thousand_cst)

        # Copied from `modeling_tf_t5.py` with -1e9 -> -10000
        if self.is_decoder and encoder_attention_mask is not None:
            # If a 2D ou 3D attention mask is provided for the cross-attention
            # we need to make broadcastable to [batch_size, num_heads, mask_seq_length, mask_seq_length]
            # we need to make broadcastable to [batch_size, num_heads, seq_length, seq_length]
            encoder_attention_mask = tf.cast(encoder_attention_mask, dtype=extended_attention_mask.dtype)
            num_dims_encoder_attention_mask = len(shape_list(encoder_attention_mask))
            if num_dims_encoder_attention_mask == 3:
                encoder_extended_attention_mask = encoder_attention_mask[:, None, :, :]
            if num_dims_encoder_attention_mask == 2:
                encoder_extended_attention_mask = encoder_attention_mask[:, None, None, :]

            # T5 has a mask that can compare sequence ids, we can simulate this here with this transposition
            # Cf. https://github.com/tensorflow/mesh/blob/8d2465e9bc93129b913b5ccc6a59aa97abd96ec6/mesh_tensorflow/transformer/transformer_layers.py#L270
            # encoder_extended_attention_mask = tf.math.equal(encoder_extended_attention_mask,
            #                                         tf.transpose(encoder_extended_attention_mask, perm=(-1, -2)))

            encoder_extended_attention_mask = (1.0 - encoder_extended_attention_mask) * -10000.0
        else:
            encoder_extended_attention_mask = None

        # Prepare head mask if needed
        # 1.0 in head_mask indicate we keep the head
        # attention_probs has shape bsz x n_heads x N x N
        # input head_mask has shape [num_heads] or [num_hidden_layers x num_heads]
        # and head_mask is converted to shape [num_hidden_layers x batch x num_heads x seq_length x seq_length]
        if head_mask is not None:
            raise NotImplementedError
        else:
            head_mask = [None] * self.config.num_hidden_layers

        encoder_outputs = self.encoder(
            hidden_states=embedding_output,
            attention_mask=extended_attention_mask,
            head_mask=head_mask,
            encoder_hidden_states=encoder_hidden_states,
            encoder_attention_mask=encoder_extended_attention_mask,
            past_key_values=past_key_values,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            training=training,
        )

        sequence_output = encoder_outputs[0]
        pooled_output = self.pooler(hidden_states=sequence_output) if self.pooler is not None else None

        if not return_dict:
            return (
                sequence_output,
                pooled_output,
            ) + encoder_outputs[1:]

        return TFBaseModelOutputWithPoolingAndCrossAttentions(
            last_hidden_state=sequence_output,
            pooler_output=pooled_output,
            past_key_values=encoder_outputs.past_key_values,
            hidden_states=encoder_outputs.hidden_states,
            attentions=encoder_outputs.attentions,
            cross_attentions=encoder_outputs.cross_attentions,
        )

    def predict_contacts(self, tokens, attention_mask):
        attns = self(tokens, attention_mask=attention_mask, return_dict=True, output_attentions=True).attentions
        attns = tf.stack(attns, axis=1)  # Matches the original model layout
        # In the original model, attentions for padding tokens are completely zeroed out.
        # This makes no difference most of the time because the other tokens won't attend to them,
        # but it does for the contact prediction task, which takes attentions as input,
        # so we have to mimic that here.
        attention_mask = tf.cast(attention_mask, attns.dtype)
        attns *= attention_mask[:, None, None, None]
        attns *= attention_mask[:, None, None, :, None]
        return self.contact_head(tokens, attns)


@add_start_docstrings(
    "The bare ESM Model transformer outputting raw hidden-states without any specific head on top.",
    ESM_START_DOCSTRING,
)
class TFEsmModel(TFEsmPreTrainedModel):
    def __init__(self, config: EsmConfig, add_pooling_layer=True, *inputs, **kwargs):
        super().__init__(config, *inputs, **kwargs)

        self.esm = TFEsmMainLayer(config, add_pooling_layer=add_pooling_layer, name="esm")

    @unpack_inputs
    @add_start_docstrings_to_model_forward(ESM_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=TFBaseModelOutputWithPoolingAndCrossAttentions,
        config_class=_CONFIG_FOR_DOC,
    )
    def call(
        self,
        input_ids: TFModelInputType | None = None,
        attention_mask: np.ndarray | tf.Tensor | None = None,
        position_ids: np.ndarray | tf.Tensor | None = None,
        head_mask: np.ndarray | tf.Tensor | None = None,
        inputs_embeds: np.ndarray | tf.Tensor | None = None,
        encoder_hidden_states: np.ndarray | tf.Tensor | None = None,
        encoder_attention_mask: np.ndarray | tf.Tensor | None = None,
        past_key_values: Optional[Tuple[Tuple[Union[np.ndarray, tf.Tensor]]]] = None,
        use_cache: Optional[bool] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        training: Optional[bool] = False,
    ) -> Union[TFBaseModelOutputWithPoolingAndCrossAttentions, Tuple[tf.Tensor]]:
        r"""
        encoder_hidden_states  (`tf.Tensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
            Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention if
            the model is configured as a decoder.
        encoder_attention_mask (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to avoid performing attention on the padding token indices of the encoder input. This mask is used in
            the cross-attention if the model is configured as a decoder. Mask values selected in `[0, 1]`:

            - 1 for tokens that are **not masked**,
            - 0 for tokens that are **masked**.

        past_key_values (`Tuple[Tuple[tf.Tensor]]` of length `config.n_layers`)
            contains precomputed key and value hidden states of the attention blocks. Can be used to speed up decoding.
            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
        use_cache (`bool`, *optional*, defaults to `True`):
            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
            `past_key_values`). Set to `False` during training, `True` during generation
        """
        outputs = self.esm(
            input_ids=input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            encoder_hidden_states=encoder_hidden_states,
            encoder_attention_mask=encoder_attention_mask,
            past_key_values=past_key_values,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            training=training,
        )
        return outputs

    def predict_contacts(self, tokens, attention_mask):
        return self.esm.predict_contacts(tokens, attention_mask)


@add_start_docstrings("""ESM Model with a `language modeling` head on top.""", ESM_START_DOCSTRING)
class TFEsmForMaskedLM(TFEsmPreTrainedModel, TFMaskedLanguageModelingLoss):
    _keys_to_ignore_on_load_missing = [r"position_ids"]
    _keys_to_ignore_on_load_unexpected = [r"pooler"]

    def __init__(self, config):
        super().__init__(config)

        if config.is_decoder:
            logger.warning(
                "If you want to use `EsmForMaskedLM` make sure `config.is_decoder=False` for "
                "bi-directional self-attention."
            )

        self.esm = TFEsmMainLayer(config, add_pooling_layer=False, name="esm")
        self.lm_head = TFEsmLMHead(config, name="lm_head")
        if config.tie_word_embeddings:
            # Ensure word embeddings are built so that we actually have something to tie
            with tf.name_scope(os.path.join(self._name_scope(), "esm", "embeddings", "word_embeddings")):
                self.esm.embeddings.word_embeddings.build((None, None))
            self.lm_head.decoder = self.esm.embeddings.word_embeddings.weights[0]

    def get_output_embeddings(self):
        return self.lm_head.decoder

    def set_output_embeddings(self, new_embeddings):
        self.lm_head.decoder = new_embeddings

    def get_lm_head(self):
        return self.lm_head

    @unpack_inputs
    @add_start_docstrings_to_model_forward(ESM_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=TFMaskedLMOutput,
        config_class=_CONFIG_FOR_DOC,
        mask="<mask>",
    )
    def call(
        self,
        input_ids: TFModelInputType | None = None,
        attention_mask: np.ndarray | tf.Tensor | None = None,
        position_ids: np.ndarray | tf.Tensor | None = None,
        head_mask: np.ndarray | tf.Tensor | None = None,
        inputs_embeds: np.ndarray | tf.Tensor | None = None,
        encoder_hidden_states: np.ndarray | tf.Tensor | None = None,
        encoder_attention_mask: np.ndarray | tf.Tensor | None = None,
        labels: np.ndarray | tf.Tensor | None = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        training: bool = False,
    ) -> Union[TFMaskedLMOutput, Tuple[tf.Tensor]]:
        r"""
        labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the masked language modeling loss. Indices should be in `[-100, 0, ...,
            config.vocab_size]` (see `input_ids` docstring) Tokens with indices set to `-100` are ignored (masked), the
            loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`
        kwargs (`Dict[str, any]`, optional, defaults to *{}*):
            Used to hide legacy arguments that have been deprecated.
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.esm(
            input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            encoder_hidden_states=encoder_hidden_states,
            encoder_attention_mask=encoder_attention_mask,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            training=training,
        )
        sequence_output = outputs[0]
        prediction_scores = self.lm_head(sequence_output)

        masked_lm_loss = None
        if labels is not None:
            masked_lm_loss = self.hf_compute_loss(labels=labels, logits=prediction_scores)

        if not return_dict:
            output = (prediction_scores,) + outputs[2:]
            return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output

        return TFMaskedLMOutput(
            loss=masked_lm_loss,
            logits=prediction_scores,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )

    def predict_contacts(self, tokens, attention_mask):
        return self.esm.predict_contacts(tokens, attention_mask)


class TFEsmLMHead(Layer):
    """ESM Head for masked language modeling."""

    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.dense = Dense(
            config.hidden_size, kernel_initializer=get_initializer(config.initializer_range), name="dense"
        )

        self.layer_norm = LayerNormalization(epsilon=config.layer_norm_eps, name="layer_norm")
        if config.tie_word_embeddings:
            self.decoder = None
        else:
            self.decoder = Dense(
                config.vocab_size,
                kernel_initializer=get_initializer(config.initializer_range),
                name="decoder",
                use_bias=False,
            )
        self.config = config

    def build(self, input_shape):
        super().build(input_shape)
        # Separate bias to match the PT model and allow weight cross-loading to work
        # Put it in the build so it gets the right name when adding it as a weight
        self.bias = self.add_weight("bias", shape=(self.config.vocab_size,), initializer="zeros", trainable=True)

    def get_bias(self):
        return {"bias": self.bias}

    def call(self, features):
        x = self.dense(features)
        x = gelu(x)
        x = self.layer_norm(x)

        # project back to size of vocabulary with bias
        if self.config.tie_word_embeddings:
            x = tf.matmul(x, self.decoder, transpose_b=True) + self.bias
        else:
            x = self.decoder(x) + self.bias
        return x


@add_start_docstrings(
    """
    ESM Model transformer with a sequence classification/regression head on top (a linear layer on top of the pooled
    output) e.g. for GLUE tasks.
    """,
    ESM_START_DOCSTRING,
)
class TFEsmForSequenceClassification(TFEsmPreTrainedModel, TFSequenceClassificationLoss):
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels
        self.config = config

        self.esm = TFEsmMainLayer(config, add_pooling_layer=False, name="esm")
        self.classifier = TFEsmClassificationHead(config, name="classifier")

    @unpack_inputs
    @add_start_docstrings_to_model_forward(ESM_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=TFSequenceClassifierOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def call(
        self,
        input_ids: TFModelInputType | None = None,
        attention_mask: np.ndarray | tf.Tensor | None = None,
        position_ids: np.ndarray | tf.Tensor | None = None,
        head_mask: np.ndarray | tf.Tensor | None = None,
        inputs_embeds: np.ndarray | tf.Tensor | None = None,
        labels: np.ndarray | tf.Tensor | None = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        training: bool = False,
    ) -> Union[TFSequenceClassifierOutput, Tuple[tf.Tensor]]:
        r"""
        labels (`tf.Tensor` of shape `(batch_size,)`, *optional*):
            Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
            config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
            `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.esm(
            input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            training=training,
        )
        sequence_output = outputs[0]
        logits = self.classifier(sequence_output)

        loss = None if labels is None else self.hf_compute_loss(labels, logits)

        if not return_dict:
            output = (logits,) + outputs[2:]
            return ((loss,) + output) if loss is not None else output

        return TFSequenceClassifierOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )


@add_start_docstrings(
    """
    ESM Model with a token classification head on top (a linear layer on top of the hidden-states output) e.g. for
    Named-Entity-Recognition (NER) tasks.
    """,
    ESM_START_DOCSTRING,
)
class TFEsmForTokenClassification(TFEsmPreTrainedModel, TFTokenClassificationLoss):
    _keys_to_ignore_on_load_unexpected = [r"pooler"]
    _keys_to_ignore_on_load_missing = [r"position_ids"]

    def __init__(self, config):
        super().__init__(config)
        self.num_labels = config.num_labels

        self.esm = TFEsmMainLayer(config, add_pooling_layer=False, name="esm")
        self.dropout = Dropout(config.hidden_dropout_prob)
        self.classifier = Dense(config.num_labels, name="classifier")

    @unpack_inputs
    @add_start_docstrings_to_model_forward(ESM_INPUTS_DOCSTRING.format("batch_size, sequence_length"))
    @add_code_sample_docstrings(
        checkpoint=_CHECKPOINT_FOR_DOC,
        output_type=TFTokenClassifierOutput,
        config_class=_CONFIG_FOR_DOC,
    )
    def call(
        self,
        input_ids: TFModelInputType | None = None,
        attention_mask: np.ndarray | tf.Tensor | None = None,
        position_ids: np.ndarray | tf.Tensor | None = None,
        head_mask: np.ndarray | tf.Tensor | None = None,
        inputs_embeds: np.ndarray | tf.Tensor | None = None,
        labels: np.ndarray | tf.Tensor | None = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        training: bool = False,
    ) -> Union[TFTokenClassifierOutput, Tuple[tf.Tensor]]:
        r"""
        labels (`tf.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
            Labels for computing the token classification loss. Indices should be in `[0, ..., config.num_labels - 1]`.
        """
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict

        outputs = self.esm(
            input_ids,
            attention_mask=attention_mask,
            position_ids=position_ids,
            head_mask=head_mask,
            inputs_embeds=inputs_embeds,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            training=training,
        )

        sequence_output = outputs[0]

        sequence_output = self.dropout(sequence_output, training=training)
        logits = self.classifier(sequence_output)

        loss = None if labels is None else self.hf_compute_loss(labels, logits)

        if not return_dict:
            output = (logits,) + outputs[2:]
            return ((loss,) + output) if loss is not None else output

        return TFTokenClassifierOutput(
            loss=loss,
            logits=logits,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
        )


class TFEsmClassificationHead(Layer):
    """Head for sentence-level classification tasks."""

    def __init__(self, config, name=None):
        super().__init__(name=name)
        self.dense = Dense(
            config.hidden_size,
            kernel_initializer=get_initializer(config.initializer_range),
            activation="tanh",
            name="dense",
        )
        self.dropout = Dropout(config.hidden_dropout_prob)
        self.out_proj = Dense(
            config.num_labels,
            kernel_initializer=get_initializer(config.initializer_range),
            activation="linear",
            name="out_proj",
        )

    def call(self, features, training=False):
        x = features[:, 0, :]  # take <s> token (equiv. to [CLS])
        x = self.dropout(x, training=training)
        x = self.dense(x)
        x = self.dropout(x, training=training)
        x = self.out_proj(x)
        return x


def create_position_ids_from_input_ids(input_ids, padding_idx, past_key_values_length=0):
    """
    Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols
    are ignored. This is modified from fairseq's `utils.make_positions`.

    Args:
        x: tf.Tensor x:

    Returns: tf.Tensor
    """
    # The series of casts and type-conversions here are carefully balanced to both work with ONNX export and XLA.
    mask = tf.cast(input_ids != padding_idx, tf.int64)
    incremental_indices = (tf.cumsum(mask, axis=1) + past_key_values_length) * mask
    return incremental_indices + padding_idx