oups

src/transformers/models/cohere/diff_cohere.py

@@ -0,0 +1,104 @@
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from transformers import CohereConfig
+from transformers.utils import ModelConverter
+
+CohereConverter = ModelConverter(__file__)
+# now should the cohere converted be added to all model converters? 
+
+class CohereLayerNorm(nn.Module):
+    def __init__(self, hidden_size=None, eps=1e-5, bias=False):
+        """The hidden size can be a tuple or an int. The tuple is used for QKNorm to normalize across head_dim"""
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(hidden_size))
+        self.variance_epsilon = eps
+
+    def forward(self, hidden_states):
+        input_dtype = hidden_states.dtype
+        hidden_states = hidden_states.to(torch.float32)
+        mean = hidden_states.mean(-1, keepdim=True)
+        variance = (hidden_states - mean).pow(2).mean(-1, keepdim=True)
+        hidden_states = (hidden_states - mean) * torch.rsqrt(variance + self.variance_epsilon)
+        hidden_states = self.weight.to(torch.float32) * hidden_states
+        return hidden_states.to(input_dtype)
+
+class CohereRotaryEmbedding(LlamaRotaryEmbedding):
+
+    def rotate_half(self, x):
+        # Split and rotate
+        x1 = x[..., ::2]
+        x2 = x[..., 1::2]
+        rot_x = torch.stack([-x2, x1], dim=-1).flatten(-2)
+        return rot_x
+
+    def forward(self, q, k, position_ids=None, unsqueeze_dim=1):
+        dtype = q.dtype
+        q,k  = q.float(), k.float()
+        cos, sin = self.comput_cos_sin(q, position_ids)
+        cos = cos.unsqueeze(unsqueeze_dim)
+        sin = sin.unsqueeze(unsqueeze_dim)
+        q_embed = (q * cos) + (self.rotate_half(q) * sin)
+        k_embed = (k * cos) + (self.rotate_half(k) * sin)
+        return q_embed.to(dtype=dtype), k_embed.to(dtype=dtype)
+
+CohereMLP = CohereConverter.register("CohereMLP", LlamaMLP) 
+CohereAttention = CohereConverter.register("CohereAttention", LlamaAttention) 
+CohereSdpaAttention = CohereConverter.register("CohereSdpaAttention", LlamaAttention) 
+CohereFlashAttention2 = CohereConverter.register("CohereFlashAttention2", LlamaAttention) 
+
+COHERE_ATTENTION_CLASSES = {"eager": CohereAttention, "flash_attention_2": CohereFlashAttention2, "sdpa": CohereSdpaAttention}
+
+class CohereDecoderLayer(nn.Module):
+    def __init__(self, config: CohereConfig, layer_idx: int):
+        super().__init__()
+        self.hidden_size = config.hidden_size
+
+        self.self_attn = COHERE_ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
+
+        self.mlp = CohereMLP(config)
+        self.input_layernorm = CohereLayerNorm(hidden_size=(config.hidden_size), eps=config.layer_norm_eps)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]:
+        residual = hidden_states
+
+        hidden_states = self.input_layernorm(hidden_states)
+
+        # Self Attention
+        hidden_states_attention, self_attn_weights, present_key_value = self.self_attn(
+            hidden_states=hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_value=past_key_value,
+            output_attentions=output_attentions,
+            use_cache=use_cache,
+            cache_position=cache_position,
+        )
+
+        # Fully Connected
+        hidden_states_mlp = self.mlp(hidden_states)
+
+        # Add everything together (main diff with llama )
+        hidden_states = residual + hidden_states_attention + hidden_states_mlp
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (self_attn_weights,)
+
+        if use_cache:
+            outputs += (present_key_value,)
+
+        return outputs
+
+CoherePreTrainedModel = CohereConverter.register("CoherePreTrainedModel", LlamaPreTrainedModel)
+CohereModel = CohereConverter.register("CohereModel", LlamaModel)
+CohereForCausalLM = CohereConverter.register("CohereForCausalLM", LlamaForCausalLM)

src/transformers/models/gemma/diff_gemma.py

@@ -0,0 +1,291 @@
+# coding=utf-8
+# Copyright 2024 Google Inc. 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 Gemma model."""
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from transformers.utils import ModelConverter
+
+
+import math
+from typing import List, Optional, Tuple, Union
+
+import torch
+import torch.utils.checkpoint
+from torch import nn
+
+from ...activations import ACT2FN
+from ...cache_utils import Cache
+
+from ...pytorch_utils import ALL_LAYERNORM_LAYERS 
+from ...utils import (
+    logging,
+)
+from .configuration_gemma import GemmaConfig
+
+from transformers.models.llama.modeling_llama import repeat_kv, apply_rotary_pos_emb
+
+
+logger = logging.get_logger(__name__)
+
+GemmaConverter = ModelConverter(__file__)
+
+class GemmaRMSNorm(nn.Module):
+    def __init__(self, dim: int, eps: float = 1e-6):
+        super().__init__()
+        self.eps = eps
+        self.weight = nn.Parameter(torch.zeros(dim))
+
+    def _norm(self, x):
+        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
+
+    def forward(self, x):
+        output = self._norm(x.float())
+        # Llama does x.to(float16) * w whilst Gemma is (x * w).to(float16)
+        # See https://github.com/huggingface/transformers/pull/29402
+        output = output * (1.0 + self.weight.float())
+        return output.type_as(x)
+
+
+ALL_LAYERNORM_LAYERS.append(GemmaRMSNorm)
+
+
+class GemmaRotaryEmbedding(nn.Module):
+    def __init__(self, dim, max_position_embeddings=2048, base=10000, device=None):
+        super().__init__()
+
+        self.dim = dim
+        self.max_position_embeddings = max_position_embeddings
+        self.base = base
+        self.register_buffer("inv_freq", None, persistent=False)
+
+    @torch.no_grad()
+    def forward(self, x, position_ids, seq_len=None):
+        # x: [bs, num_attention_heads, seq_len, head_size]
+        if self.inv_freq is None:
+            self.inv_freq = 1.0 / (
+                self.base ** (torch.arange(0, self.dim, 2, dtype=torch.int64, device=x.device).float() / self.dim)
+            )
+        inv_freq_expanded = self.inv_freq[None, :, None].float().expand(position_ids.shape[0], -1, 1)
+        position_ids_expanded = position_ids[:, None, :].float()
+        # Force float32 since bfloat16 loses precision on long contexts
+        # See https://github.com/huggingface/transformers/pull/29285
+        device_type = x.device.type
+        device_type = device_type if isinstance(device_type, str) and device_type != "mps" else "cpu"
+        with torch.autocast(device_type=device_type, enabled=False):
+            freqs = (inv_freq_expanded.float() @ position_ids_expanded.float()).transpose(1, 2)
+            emb = torch.cat((freqs, freqs), dim=-1)
+            cos = emb.cos()
+            sin = emb.sin()
+        return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
+
+
+
+class GemmaMLP(nn.Module):
+    def __init__(self, config):
+        super().__init__()
+        self.config = config
+        self.hidden_size = config.hidden_size
+        self.intermediate_size = config.intermediate_size
+        self.gate_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False)
+        self.up_proj = nn.Linear(self.hidden_size, self.intermediate_size, bias=False)
+        self.down_proj = nn.Linear(self.intermediate_size, self.hidden_size, bias=False)
+        if config.hidden_activation is None:
+            logger.warning_once(
+                "`config.hidden_act` is ignored, you should use `config.hidden_activation` instead.\n"
+                "Gemma's activation function will be set to `gelu_pytorch_tanh`. Please, use\n"
+                "`config.hidden_activation` if you want to override this behaviour.\n"
+                "See https://github.com/huggingface/transformers/pull/29402 for more details."
+            )
+            config.hidden_activation = "gelu_pytorch_tanh"
+        hidden_activation = config.hidden_activation
+        self.act_fn = ACT2FN[hidden_activation]
+
+    def forward(self, x):
+        return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
+
+
+
+class GemmaAttention(nn.Module):
+    """Multi-headed attention from 'Attention Is All You Need' paper"""
+
+    # Ignore copy
+    def __init__(self, config: GemmaConfig, layer_idx: Optional[int] = None):
+        super().__init__()
+        self.config = config
+        self.layer_idx = layer_idx
+        if layer_idx is None:
+            logger.warning_once(
+                f"Instantiating {self.__class__.__name__} without passing a `layer_idx` is not recommended and will "
+                "lead to errors during the forward call if caching is used. Please make sure to provide a `layer_idx` "
+                "when creating this class."
+            )
+
+        self.attention_dropout = config.attention_dropout
+        self.hidden_size = config.hidden_size
+        self.num_heads = config.num_attention_heads
+        self.head_dim = config.head_dim
+        self.num_key_value_heads = config.num_key_value_heads
+        self.num_key_value_groups = self.num_heads // self.num_key_value_heads
+        self.max_position_embeddings = config.max_position_embeddings
+        self.rope_theta = config.rope_theta
+        self.is_causal = True
+
+        if self.hidden_size % self.num_heads != 0:
+            raise ValueError(
+                f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}"
+                f" and `num_heads`: {self.num_heads})."
+            )
+
+        self.q_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.attention_bias)
+        self.k_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=config.attention_bias)
+        self.v_proj = nn.Linear(self.hidden_size, self.num_key_value_heads * self.head_dim, bias=config.attention_bias)
+        self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=config.attention_bias)
+        self.rotary_emb = GemmaRotaryEmbedding(
+            self.head_dim,
+            max_position_embeddings=self.max_position_embeddings,
+            base=self.rope_theta,
+        )
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        cache_position: Optional[torch.LongTensor] = None,
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        bsz, q_len, _ = hidden_states.size()
+
+        query_states = self.q_proj(hidden_states)
+        key_states = self.k_proj(hidden_states)
+        value_states = self.v_proj(hidden_states)
+
+        query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
+        key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+        value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+
+        cos, sin = self.rotary_emb(value_states, position_ids, seq_len=None)
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, None)
+
+        if past_key_value is not None:
+            # sin and cos are specific to RoPE models; cache_position needed for the static cache
+            cache_kwargs = {"sin": sin, "cos": cos, "cache_position": cache_position}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
+
+        if attention_mask is not None:  # no matter the length, we just slice it
+            causal_mask = attention_mask[:, :, :, : key_states.shape[-2]]
+            attn_weights = attn_weights + causal_mask
+
+        # upcast attention to fp32
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query_states.dtype)
+        attn_weights = nn.functional.dropout(attn_weights, p=self.attention_dropout, training=self.training)
+        attn_output = torch.matmul(attn_weights, value_states)
+
+        if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
+            raise ValueError(
+                f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is"
+                f" {attn_output.size()}"
+            )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+
+        attn_output = attn_output.view(bsz, q_len, -1)
+        attn_output = self.o_proj(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+
+GemmaFlashAttention2 = GemmaConverter.register("GemmaFlashAttention2", LlamaFlashAttention2)
+GemmaSdpaAttention = GemmaConverter.register("GemmaSdpaAttention", LlamaSdpaAttention)
+
+COHERE_ATTENTION_CLASSES = {"eager": GemmaAttention, "flash_attention_2": GemmaFlashAttention2, "sdpa": GemmaSdpaAttention}
+
+GemmaConverter.register("GemmaDecoderLayer", LlamaDecoderLayer) 
+GemmaConverter.register("GemmaPreTrainedModel", LlamaPreTrainedModel)
+
+class GemmaModel(LlamaModel):
+    
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        output_attentions = output_attentions | self.config.output_attentions
+        output_hidden_states = output_hidden_states| self.config.output_hidden_states
+        return_dict = return_dict | self.config.use_return_dict
+
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        logits = logits.float()
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            shift_logits = shift_logits.view(-1, self.config.vocab_size)
+            shift_labels = shift_labels.view(-1)
+            # Enable model parallelism
+            shift_labels = shift_labels.to(shift_logits.device)
+            loss = loss_fct(shift_logits, shift_labels)
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
+
+GemmaConverter.register("GemmaForCausalLM", LlamaForCausalLM)

src/transformers/models/gemma/modeling_gemma.py

@@ -1,3 +1,14 @@
+# 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.
 # coding=utf-8
 # Copyright 2024 Google Inc. HuggingFace Inc. team. All rights reserved.
 #
@@ -14,68 +25,32 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 """ PyTorch Gemma model."""
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from transformers.utils import ModelConverter
+
 
 import math
-import warnings
 from typing import List, Optional, Tuple, Union
 
 import torch
-import torch.nn.functional as F
 import torch.utils.checkpoint
 from torch import nn
-from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
 
 from ...activations import ACT2FN
-from ...cache_utils import Cache, DynamicCache, StaticCache
-from ...modeling_attn_mask_utils import (
-    AttentionMaskConverter,
-    _prepare_4d_causal_attention_mask,
-)
-from ...modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast, SequenceClassifierOutputWithPast
-from ...modeling_utils import PreTrainedModel
-from ...pytorch_utils import ALL_LAYERNORM_LAYERS, is_torch_greater_or_equal_than_1_13
+from ...cache_utils import Cache
+
+from ...pytorch_utils import ALL_LAYERNORM_LAYERS 
 from ...utils import (
-    add_start_docstrings,
-    add_start_docstrings_to_model_forward,
-    is_flash_attn_2_available,
-    is_flash_attn_greater_or_equal_2_10,
     logging,
-    replace_return_docstrings,
 )
-from ...utils.import_utils import is_torch_fx_available
 from .configuration_gemma import GemmaConfig
 
-
-if is_flash_attn_2_available():
-    from flash_attn import flash_attn_func, flash_attn_varlen_func
-    from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input  # noqa
-
-
-# This makes `_prepare_4d_causal_attention_mask` a leaf function in the FX graph.
-# It means that the function will not be traced through and simply appear as a node in the graph.
-if is_torch_fx_available():
-    if not is_torch_greater_or_equal_than_1_13:
-        import torch.fx
-
-    _prepare_4d_causal_attention_mask = torch.fx.wrap(_prepare_4d_causal_attention_mask)
+from transformers.models.llama.modeling_llama import repeat_kv, apply_rotary_pos_emb
 
 
 logger = logging.get_logger(__name__)
 
-_CONFIG_FOR_DOC = "GemmaConfig"
-
-
-def _get_unpad_data(attention_mask):
-    seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32)
-    indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten()
-    max_seqlen_in_batch = seqlens_in_batch.max().item()
-    cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.torch.int32), (1, 0))
-    return (
-        indices,
-        cu_seqlens,
-        max_seqlen_in_batch,
-    )
-
 
 class GemmaRMSNorm(nn.Module):
     def __init__(self, dim: int, eps: float = 1e-6):
@@ -127,41 +102,6 @@ class GemmaRotaryEmbedding(nn.Module):
         return cos.to(dtype=x.dtype), sin.to(dtype=x.dtype)
 
 
-# Copied from transformers.models.llama.modeling_llama.rotate_half
-def rotate_half(x):
-    """Rotates half the hidden dims of the input."""
-    x1 = x[..., : x.shape[-1] // 2]
-    x2 = x[..., x.shape[-1] // 2 :]
-    return torch.cat((-x2, x1), dim=-1)
-
-
-# Copied from transformers.models.llama.modeling_llama.apply_rotary_pos_emb
-def apply_rotary_pos_emb(q, k, cos, sin, position_ids=None, unsqueeze_dim=1):
-    """Applies Rotary Position Embedding to the query and key tensors.
-
-    Args:
-        q (`torch.Tensor`): The query tensor.
-        k (`torch.Tensor`): The key tensor.
-        cos (`torch.Tensor`): The cosine part of the rotary embedding.
-        sin (`torch.Tensor`): The sine part of the rotary embedding.
-        position_ids (`torch.Tensor`, *optional*):
-            Deprecated and unused.
-        unsqueeze_dim (`int`, *optional*, defaults to 1):
-            The 'unsqueeze_dim' argument specifies the dimension along which to unsqueeze cos[position_ids] and
-            sin[position_ids] so that they can be properly broadcasted to the dimensions of q and k. For example, note
-            that cos[position_ids] and sin[position_ids] have the shape [batch_size, seq_len, head_dim]. Then, if q and
-            k have the shape [batch_size, heads, seq_len, head_dim], then setting unsqueeze_dim=1 makes
-            cos[position_ids] and sin[position_ids] broadcastable to the shapes of q and k. Similarly, if q and k have
-            the shape [batch_size, seq_len, heads, head_dim], then set unsqueeze_dim=2.
-    Returns:
-        `tuple(torch.Tensor)` comprising of the query and key tensors rotated using the Rotary Position Embedding.
-    """
-    cos = cos.unsqueeze(unsqueeze_dim)
-    sin = sin.unsqueeze(unsqueeze_dim)
-    q_embed = (q * cos) + (rotate_half(q) * sin)
-    k_embed = (k * cos) + (rotate_half(k) * sin)
-    return q_embed, k_embed
-
 
 class GemmaMLP(nn.Module):
     def __init__(self, config):
@@ -187,18 +127,6 @@ class GemmaMLP(nn.Module):
         return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
 
 
-# Copied from transformers.models.llama.modeling_llama.repeat_kv
-def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:
-    """
-    This is the equivalent of torch.repeat_interleave(x, dim=1, repeats=n_rep). The hidden states go from (batch,
-    num_key_value_heads, seqlen, head_dim) to (batch, num_attention_heads, seqlen, head_dim)
-    """
-    batch, num_key_value_heads, slen, head_dim = hidden_states.shape
-    if n_rep == 1:
-        return hidden_states
-    hidden_states = hidden_states[:, :, None, :, :].expand(batch, num_key_value_heads, n_rep, slen, head_dim)
-    return hidden_states.reshape(batch, num_key_value_heads * n_rep, slen, head_dim)
-
 
 class GemmaAttention(nn.Module):
     """Multi-headed attention from 'Attention Is All You Need' paper"""
@@ -301,7 +229,6 @@ class GemmaAttention(nn.Module):
         return attn_output, attn_weights, past_key_value
 
 
-# Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2 with Llama->Gemma
 class GemmaFlashAttention2(GemmaAttention):
     """
     Gemma flash attention module. This module inherits from `GemmaAttention` as the weights of the module stays
@@ -317,7 +244,6 @@ class GemmaFlashAttention2(GemmaAttention):
         # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left).
         self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10()
 
-    # Ignore copy
     def forward(
         self,
         hidden_states: torch.Tensor,
@@ -334,6 +260,7 @@ class GemmaFlashAttention2(GemmaAttention):
                 "`static` cache implementation is not compatible with `attn_implementation==flash_attention_2` "
                 "make sure to use `sdpa` in the mean time, and open an issue at https://github.com/huggingface/transformers"
             )
+
         output_attentions = False
 
         bsz, q_len, _ = hidden_states.size()
@@ -349,8 +276,8 @@ class GemmaFlashAttention2(GemmaAttention):
         key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
         value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
 
-        cos, sin = self.rotary_emb(value_states, position_ids, seq_len=None)
-        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, None)
+        cos, sin = self.rotary_emb(value_states, position_ids)
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
 
         if past_key_value is not None:
             # sin and cos are specific to RoPE models; cache_position needed for the static cache
@@ -395,7 +322,7 @@ class GemmaFlashAttention2(GemmaAttention):
             query_states, key_states, value_states, attention_mask, q_len, dropout=dropout_rate
         )
 
-        attn_output = attn_output.reshape(bsz, q_len, -1).contiguous()
+        attn_output = attn_output.reshape(bsz, q_len, self.hidden_size).contiguous()
         attn_output = self.o_proj(attn_output)
 
         if not output_attentions:
@@ -500,8 +427,6 @@ class GemmaFlashAttention2(GemmaAttention):
             (max_seqlen_in_batch_q, max_seqlen_in_batch_k),
         )
 
-
-# Copied from transformers.models.llama.modeling_llama.LlamaSdpaAttention with Llama->Gemma
 class GemmaSdpaAttention(GemmaAttention):
     """
     Gemma attention module using torch.nn.functional.scaled_dot_product_attention. This module inherits from
@@ -509,7 +434,7 @@ class GemmaSdpaAttention(GemmaAttention):
     SDPA API.
     """
 
-    # Ignore copy
+    # Adapted from GemmaAttention.forward
     def forward(
         self,
         hidden_states: torch.Tensor,
@@ -546,8 +471,8 @@ class GemmaSdpaAttention(GemmaAttention):
         key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
         value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
 
-        cos, sin = self.rotary_emb(value_states, position_ids, seq_len=None)
-        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, None)
+        cos, sin = self.rotary_emb(value_states, position_ids)
+        query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin)
 
         if past_key_value is not None:
             # sin and cos are specific to RoPE models; cache_position needed for the static cache
@@ -582,27 +507,21 @@ class GemmaSdpaAttention(GemmaAttention):
         )
 
         attn_output = attn_output.transpose(1, 2).contiguous()
-        attn_output = attn_output.view(bsz, q_len, -1)
+        attn_output = attn_output.view(bsz, q_len, self.hidden_size)
 
         attn_output = self.o_proj(attn_output)
 
         return attn_output, None, past_key_value
 
 
-GEMMA_ATTENTION_CLASSES = {
-    "eager": GemmaAttention,
-    "flash_attention_2": GemmaFlashAttention2,
-    "sdpa": GemmaSdpaAttention,
-}
+COHERE_ATTENTION_CLASSES = {"eager": GemmaAttention, "flash_attention_2": GemmaFlashAttention2, "sdpa": GemmaSdpaAttention}
 
-
-# Copied from transformers.models.llama.modeling_llama.LlamaDecoderLayer with LLAMA->GEMMA,Llama->Gemma
 class GemmaDecoderLayer(nn.Module):
     def __init__(self, config: GemmaConfig, layer_idx: int):
         super().__init__()
         self.hidden_size = config.hidden_size
 
-        self.self_attn = GEMMA_ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
+        self.self_attn = LLAMA_ATTENTION_CLASSES[config._attn_implementation](config=config, layer_idx=layer_idx)
 
         self.mlp = GemmaMLP(config)
         self.input_layernorm = GemmaRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
@@ -671,35 +590,16 @@ class GemmaDecoderLayer(nn.Module):
 
         return outputs
 
-
-GEMMA_START_DOCSTRING = r"""
-    This model inherits from [`PreTrainedModel`]. 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 PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
-    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
-    and behavior.
-
-    Parameters:
-        config ([`GemmaConfig`]):
-            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
-            [`~PreTrainedModel.from_pretrained`] method to load the model weights.
-"""
-
-
 @add_start_docstrings(
-    "The bare Gemma Model outputting raw hidden-states without any specific head on top.",
-    GEMMA_START_DOCSTRING,
+    "The bare LLaMA Model outputting raw hidden-states without any specific head on top.",
+    LLAMA_START_DOCSTRING,
 )
 class GemmaPreTrainedModel(PreTrainedModel):
     config_class = GemmaConfig
     base_model_prefix = "model"
     supports_gradient_checkpointing = True
-    _keep_in_fp32_modules = ["inv_freq", "rotary_emb", "cos_cached", "sin_cached"]
     _no_split_modules = ["GemmaDecoderLayer"]
-    _skip_keys_device_placement = ["past_key_values", "causal_mask"]
+    _skip_keys_device_placement = ["past_key_values"]
     _supports_flash_attn_2 = True
     _supports_sdpa = True
     _supports_cache_class = True
@@ -716,116 +616,36 @@ class GemmaPreTrainedModel(PreTrainedModel):
                 module.weight.data[module.padding_idx].zero_()
 
 
-GEMMA_INPUTS_DOCSTRING = r"""
-    Args:
-        input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
-            Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
-            it.
-
-            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
-            [`PreTrainedTokenizer.__call__`] for details.
-
-            [What are input IDs?](../glossary#input-ids)
-        attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *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)
-
-            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
-            [`PreTrainedTokenizer.__call__`] for details.
-
-            If `past_key_values` is used, optionally only the last `input_ids` have to be input (see
-            `past_key_values`).
-
-            If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
-            and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
-            information on the default strategy.
-
-            - 1 indicates the head is **not masked**,
-            - 0 indicates the head is **masked**.
-        position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
-            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
-            config.n_positions - 1]`.
-
-            [What are position IDs?](../glossary#position-ids)
-        past_key_values (`Cache` or `tuple(tuple(torch.FloatTensor))`, *optional*):
-            Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
-            blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values`
-            returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.
-
-            Two formats are allowed:
-            - a [`~cache_utils.Cache`] instance;
-            - Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of
-            shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`). This is also known as the legacy
-            cache format.
-
-            The model will output the same cache format that is fed as input. If no `past_key_values` are passed, the
-            legacy cache format will be returned.
-
-            If `past_key_values` are used, the user can optionally input only the last `input_ids` (those that don't
-            have their past key value states given to this model) of shape `(batch_size, 1)` instead of all `input_ids`
-            of shape `(batch_size, sequence_length)`.
-        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, 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.
-        use_cache (`bool`, *optional*):
-            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
-            `past_key_values`).
-        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 [`~utils.ModelOutput`] instead of a plain tuple.
-        cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*):
-            Indices depicting the position of the input sequence tokens in the sequence. Contrarily to `position_ids`,
-            this tensor is not affected by padding. It is used to update the cache in the correct position and to infer
-            the complete sequence length.
-"""
-
-
 @add_start_docstrings(
-    "The bare Gemma Model outputting raw hidden-states without any specific head on top.",
-    GEMMA_START_DOCSTRING,
+    "The bare LLaMA Model outputting raw hidden-states without any specific head on top.",
+    LLAMA_START_DOCSTRING,
 )
-# Copied from transformers.models.llama.modeling_llama.LlamaModel with LLAMA->GEMMA,Llama->Gemma
-class GemmaModel(GemmaPreTrainedModel):
+class LlamaModel(LlamaPreTrainedModel):
     """
-    Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`GemmaDecoderLayer`]
+    Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`LlamaDecoderLayer`]
 
     Args:
-        config: GemmaConfig
+        config: LlamaConfig
     """
-
-    def __init__(self, config: GemmaConfig):
+    def __init__(self, config: LlamaConfig):
         super().__init__(config)
         self.padding_idx = config.pad_token_id
         self.vocab_size = config.vocab_size
 
         self.embed_tokens = nn.Embedding(config.vocab_size, config.hidden_size, self.padding_idx)
         self.layers = nn.ModuleList(
-            [GemmaDecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)]
+            [LlamaDecoderLayer(config, layer_idx) for layer_idx in range(config.num_hidden_layers)]
         )
-        self.norm = GemmaRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
+        self.norm = LlamaRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
         self.gradient_checkpointing = False
 
         # Initialize weights and apply final processing
         self.post_init()
-
     def get_input_embeddings(self):
         return self.embed_tokens
-
     def set_input_embeddings(self, value):
         self.embed_tokens = value
-
-    @add_start_docstrings_to_model_forward(GEMMA_INPUTS_DOCSTRING)
-    # Ignore copy
+    
     def forward(
         self,
         input_ids: torch.LongTensor = None,
@@ -833,114 +653,56 @@ class GemmaModel(GemmaPreTrainedModel):
         position_ids: Optional[torch.LongTensor] = None,
         past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
         inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
         use_cache: Optional[bool] = None,
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
         cache_position: Optional[torch.LongTensor] = None,
-    ) -> Union[Tuple, BaseModelOutputWithPast]:
-        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
-        output_hidden_states = (
-            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        output_attentions = output_attentions | self.config.output_attentions
+        output_hidden_states = output_hidden_states| self.config.output_hidden_states
+        return_dict = return_dict | self.config.use_return_dict
+
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
         )
-        use_cache = use_cache if use_cache is not None else self.config.use_cache
-        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
 
-        if (input_ids is None) ^ (inputs_embeds is not None):
-            raise ValueError(
-                "You cannot specify both input_ids and inputs_embeds at the same time, and must specify either one"
-            )
-
-        if self.gradient_checkpointing and self.training and use_cache:
-            logger.warning_once(
-                "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`."
-            )
-            use_cache = False
-
-        if inputs_embeds is None:
-            inputs_embeds = self.embed_tokens(input_ids)
-
-        return_legacy_cache = False
-        if use_cache and not isinstance(past_key_values, Cache):  # kept for BC (non `Cache` `past_key_values` inputs)
-            return_legacy_cache = True
-            past_key_values = DynamicCache.from_legacy_cache(past_key_values)
-
-        if cache_position is None:
-            past_seen_tokens = past_key_values.get_seq_length() if past_key_values is not None else 0
-            cache_position = torch.arange(
-                past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device
-            )
-
-        if position_ids is None:
-            position_ids = cache_position.unsqueeze(0)
-
-        causal_mask = self._update_causal_mask(attention_mask, inputs_embeds, cache_position, past_key_values)
-
-        # embed positions
-        hidden_states = inputs_embeds
-
-        # normalized
-        # Gemma downcasts the below to float16, causing sqrt(3072)=55.4256 to become 55.5
-        # See https://github.com/huggingface/transformers/pull/29402
-        normalizer = torch.tensor(self.config.hidden_size**0.5, dtype=hidden_states.dtype)
-        hidden_states = hidden_states * normalizer
-
-        # decoder layers
-        all_hidden_states = () if output_hidden_states else None
-        all_self_attns = () if output_attentions else None
-        next_decoder_cache = None
-
-        for decoder_layer in self.layers:
-            if output_hidden_states:
-                all_hidden_states += (hidden_states,)
-
-            if self.gradient_checkpointing and self.training:
-                layer_outputs = self._gradient_checkpointing_func(
-                    decoder_layer.__call__,
-                    hidden_states,
-                    causal_mask,
-                    position_ids,
-                    past_key_values,
-                    output_attentions,
-                    use_cache,
-                    cache_position,
-                )
-            else:
-                layer_outputs = decoder_layer(
-                    hidden_states,
-                    attention_mask=causal_mask,
-                    position_ids=position_ids,
-                    past_key_value=past_key_values,
-                    output_attentions=output_attentions,
-                    use_cache=use_cache,
-                    cache_position=cache_position,
-                )
-
-            hidden_states = layer_outputs[0]
-
-            if use_cache:
-                next_decoder_cache = layer_outputs[2 if output_attentions else 1]
-
-            if output_attentions:
-                all_self_attns += (layer_outputs[1],)
-
-        hidden_states = self.norm(hidden_states)
-
-        # add hidden states from the last decoder layer
-        if output_hidden_states:
-            all_hidden_states += (hidden_states,)
-
-        next_cache = next_decoder_cache if use_cache else None
-        if return_legacy_cache:
-            next_cache = next_cache.to_legacy_cache()
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        logits = logits.float()
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            shift_logits = shift_logits.view(-1, self.config.vocab_size)
+            shift_labels = shift_labels.view(-1)
+            # Enable model parallelism
+            shift_labels = shift_labels.to(shift_logits.device)
+            loss = loss_fct(shift_logits, shift_labels)
 
         if not return_dict:
-            return tuple(v for v in [hidden_states, next_cache, all_hidden_states, all_self_attns] if v is not None)
-        return BaseModelOutputWithPast(
-            last_hidden_state=hidden_states,
-            past_key_values=next_cache,
-            hidden_states=all_hidden_states,
-            attentions=all_self_attns,
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
         )
 
     def _update_causal_mask(
@@ -1029,8 +791,66 @@ class GemmaModel(GemmaPreTrainedModel):
 
         return causal_mask
 
+    
+    def forward(
+        self,
+        input_ids: torch.LongTensor = None,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
+        inputs_embeds: Optional[torch.FloatTensor] = None,
+        labels: Optional[torch.LongTensor] = None,
+        use_cache: Optional[bool] = None,
+        output_attentions: Optional[bool] = None,
+        output_hidden_states: Optional[bool] = None,
+        return_dict: Optional[bool] = None,
+        cache_position: Optional[torch.LongTensor] = None,
+    ) -> Union[Tuple, CausalLMOutputWithPast]:
+        output_attentions = output_attentions | self.config.output_attentions
+        output_hidden_states = output_hidden_states| self.config.output_hidden_states
+        return_dict = return_dict | self.config.use_return_dict
+
+        outputs = self.model(
+            input_ids=input_ids,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_values=past_key_values,
+            inputs_embeds=inputs_embeds,
+            use_cache=use_cache,
+            output_attentions=output_attentions,
+            output_hidden_states=output_hidden_states,
+            return_dict=return_dict,
+            cache_position=cache_position,
+        )
+
+        hidden_states = outputs[0]
+        logits = self.lm_head(hidden_states)
+        logits = logits.float()
+        loss = None
+        if labels is not None:
+            # Shift so that tokens < n predict n
+            shift_logits = logits[..., :-1, :].contiguous()
+            shift_labels = labels[..., 1:].contiguous()
+            # Flatten the tokens
+            loss_fct = CrossEntropyLoss()
+            shift_logits = shift_logits.view(-1, self.config.vocab_size)
+            shift_labels = shift_labels.view(-1)
+            # Enable model parallelism
+            shift_labels = shift_labels.to(shift_logits.device)
+            loss = loss_fct(shift_logits, shift_labels)
+
+        if not return_dict:
+            output = (logits,) + outputs[1:]
+            return (loss,) + output if loss is not None else output
+
+        return CausalLMOutputWithPast(
+            loss=loss,
+            logits=logits,
+            past_key_values=outputs.past_key_values,
+            hidden_states=outputs.hidden_states,
+            attentions=outputs.attentions,
+        )
 
-# Copied from transformers.models.llama.modeling_llama.LlamaForCausalLM with LLAMA->GEMMA,Llama->Gemma,llama->gemma
 class GemmaForCausalLM(GemmaPreTrainedModel):
     _tied_weights_keys = ["lm_head.weight"]
 
@@ -1061,8 +881,7 @@ class GemmaForCausalLM(GemmaPreTrainedModel):
     def get_decoder(self):
         return self.model
 
-    # Ignore copy
-    @add_start_docstrings_to_model_forward(GEMMA_INPUTS_DOCSTRING)
+    @add_start_docstrings_to_model_forward(LLAMA_INPUTS_DOCSTRING)
     @replace_return_docstrings(output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
     def forward(
         self,
@@ -1092,16 +911,16 @@ class GemmaForCausalLM(GemmaPreTrainedModel):
         ```python
         >>> from transformers import AutoTokenizer, GemmaForCausalLM
 
-        >>> model = GemmaForCausalLM.from_pretrained("google/gemma-7b")
-        >>> tokenizer = AutoTokenizer.from_pretrained("google/gemma-7b")
+        >>> model = GemmaForCausalLM.from_pretrained("meta-llama/Gemma-2-7b-hf")
+        >>> tokenizer = AutoTokenizer.from_pretrained("meta-llama/Gemma-2-7b-hf")
 
-        >>> prompt = "What is your favorite condiment?"
+        >>> prompt = "Hey, are you conscious? Can you talk to me?"
         >>> inputs = tokenizer(prompt, return_tensors="pt")
 
         >>> # Generate
         >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
         >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
-        "What is your favorite condiment?"
+        "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
         ```"""
         output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
         output_hidden_states = (
@@ -1124,8 +943,14 @@ class GemmaForCausalLM(GemmaPreTrainedModel):
         )
 
         hidden_states = outputs[0]
-        logits = self.lm_head(hidden_states)
+        if self.config.pretraining_tp > 1:
+            lm_head_slices = self.lm_head.weight.split(self.vocab_size // self.config.pretraining_tp, dim=0)
+            logits = [F.linear(hidden_states, lm_head_slices[i]) for i in range(self.config.pretraining_tp)]
+            logits = torch.cat(logits, dim=-1)
+        else:
+            logits = self.lm_head(hidden_states)
         logits = logits.float()
+
         loss = None
         if labels is not None:
             # Shift so that tokens < n predict n
@@ -1238,126 +1063,3 @@ class GemmaForCausalLM(GemmaPreTrainedModel):
             )
         return reordered_past
 
-
-@add_start_docstrings(
-    """
-    The Gemma Model transformer with a sequence classification head on top (linear layer).
-
-    [`GemmaForSequenceClassification`] uses the last token in order to do the classification, as other causal models
-    (e.g. GPT-2) do.
-
-    Since it does classification on the last token, it requires to know the position of the last token. If a
-    `pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If
-    no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the
-    padding tokens when `inputs_embeds` are passed instead of `input_ids`, it does the same (take the last value in
-    each row of the batch).
-    """,
-    GEMMA_START_DOCSTRING,
-)
-# Copied from transformers.models.llama.modeling_llama.LlamaForSequenceClassification with LLAMA->GEMMA,Llama->Gemma
-class GemmaForSequenceClassification(GemmaPreTrainedModel):
-    def __init__(self, config):
-        super().__init__(config)
-        self.num_labels = config.num_labels
-        self.model = GemmaModel(config)
-        self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False)
-
-        # Initialize weights and apply final processing
-        self.post_init()
-
-    def get_input_embeddings(self):
-        return self.model.embed_tokens
-
-    def set_input_embeddings(self, value):
-        self.model.embed_tokens = value
-
-    @add_start_docstrings_to_model_forward(GEMMA_INPUTS_DOCSTRING)
-    def forward(
-        self,
-        input_ids: torch.LongTensor = None,
-        attention_mask: Optional[torch.Tensor] = None,
-        position_ids: Optional[torch.LongTensor] = None,
-        past_key_values: Optional[Union[Cache, List[torch.FloatTensor]]] = None,
-        inputs_embeds: Optional[torch.FloatTensor] = None,
-        labels: Optional[torch.LongTensor] = None,
-        use_cache: Optional[bool] = None,
-        output_attentions: Optional[bool] = None,
-        output_hidden_states: Optional[bool] = None,
-        return_dict: Optional[bool] = None,
-    ) -> Union[Tuple, SequenceClassifierOutputWithPast]:
-        r"""
-        labels (`torch.LongTensor` 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
-
-        transformer_outputs = self.model(
-            input_ids,
-            attention_mask=attention_mask,
-            position_ids=position_ids,
-            past_key_values=past_key_values,
-            inputs_embeds=inputs_embeds,
-            use_cache=use_cache,
-            output_attentions=output_attentions,
-            output_hidden_states=output_hidden_states,
-            return_dict=return_dict,
-        )
-        hidden_states = transformer_outputs[0]
-        logits = self.score(hidden_states)
-
-        if input_ids is not None:
-            batch_size = input_ids.shape[0]
-        else:
-            batch_size = inputs_embeds.shape[0]
-
-        if self.config.pad_token_id is None and batch_size != 1:
-            raise ValueError("Cannot handle batch sizes > 1 if no padding token is defined.")
-        if self.config.pad_token_id is None:
-            sequence_lengths = -1
-        else:
-            if input_ids is not None:
-                # if no pad token found, use modulo instead of reverse indexing for ONNX compatibility
-                sequence_lengths = torch.eq(input_ids, self.config.pad_token_id).int().argmax(-1) - 1
-                sequence_lengths = sequence_lengths % input_ids.shape[-1]
-                sequence_lengths = sequence_lengths.to(logits.device)
-            else:
-                sequence_lengths = -1
-
-        pooled_logits = logits[torch.arange(batch_size, device=logits.device), sequence_lengths]
-
-        loss = None
-        if labels is not None:
-            labels = labels.to(logits.device)
-            if self.config.problem_type is None:
-                if self.num_labels == 1:
-                    self.config.problem_type = "regression"
-                elif self.num_labels > 1 and (labels.dtype == torch.long or labels.dtype == torch.int):
-                    self.config.problem_type = "single_label_classification"
-                else:
-                    self.config.problem_type = "multi_label_classification"
-
-            if self.config.problem_type == "regression":
-                loss_fct = MSELoss()
-                if self.num_labels == 1:
-                    loss = loss_fct(pooled_logits.squeeze(), labels.squeeze())
-                else:
-                    loss = loss_fct(pooled_logits, labels)
-            elif self.config.problem_type == "single_label_classification":
-                loss_fct = CrossEntropyLoss()
-                loss = loss_fct(pooled_logits.view(-1, self.num_labels), labels.view(-1))
-            elif self.config.problem_type == "multi_label_classification":
-                loss_fct = BCEWithLogitsLoss()
-                loss = loss_fct(pooled_logits, labels)
-        if not return_dict:
-            output = (pooled_logits,) + transformer_outputs[1:]
-            return ((loss,) + output) if loss is not None else output
-
-        return SequenceClassifierOutputWithPast(
-            loss=loss,
-            logits=pooled_logits,
-            past_key_values=transformer_outputs.past_key_values,
-            hidden_states=transformer_outputs.hidden_states,
-            attentions=transformer_outputs.attentions,
-        )

src/transformers/models/persimmon/diff_persimmon.py

@@ -0,0 +1,87 @@
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from .configuration_persimmon import PersimmonConfig
+from transformers.utils import ModelConverter
+
+PersimmonConverter = ModelConverter(__file__)
+
+PersimmonConverter.register("PersimmonRotaryEmbedding", LlamaRotaryEmbedding)
+PersimmonConverter.register("PersimmonMLP", LlamaMLP) 
+
+class PersimmonAttention(LlamaAttention):
+    """Multi-headed attention from 'Attention Is All You Need' paper"""
+
+    def __init__(self, config: PersimmonConfig, layer_idx: Optional[int] = None):
+        super().__init__()
+        ... # copy before? add the line? how to best support this
+        self.query_key_value = nn.Linear(self.hidden_size, 3 * self.hidden_size, bias=True)
+        self.dense = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=True)
+        self.qk_layernorm = config.qk_layernorm
+
+        if self.qk_layernorm:
+            self.q_layernorm = nn.LayerNorm(
+                config.hidden_size // self.num_heads, eps=config.layer_norm_eps, elementwise_affine=True
+            )
+            self.k_layernorm = nn.LayerNorm(
+                config.hidden_size // self.num_heads, eps=config.layer_norm_eps, elementwise_affine=True
+            )
+        self.attention_dropout = nn.Dropout(config.attention_dropout)
+        self._init_rope()
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        bsz, q_len, _ = hidden_states.size()
+
+        # [batch_size, seq_length, 3 x hidden_size]
+        fused_qkv = self.query_key_value(hidden_states)
+
+        # 3 x [batch_size, seq_length, num_heads, head_dim]
+        (query_states, key_states, value_states) = self._split_heads(fused_qkv)
+
+        if self.qk_layernorm:
+            query_states = self.q_layernorm(query_states)
+            key_states = self.k_layernorm(key_states)
+
+        # [batch_size, num_heads, seq_length, head_dim] -> [batch_size, seq_length, num_heads, head_dim]
+        query_states = query_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+
+        os, sin = self.rotary_emb(value_states, seq_len=None)
+
+        # Partial rotary embedding
+        query_rot, query_pass = (
+            query_states[..., : self.rotary_emb.dim],
+            query_states[..., self.rotary_emb.dim :],
+        )
+        key_rot, key_pass = (
+            key_states[..., : self.rotary_emb.dim],
+            key_states[..., self.rotary_emb.dim :],
+        )
+        # [batch_size, seq_length, num_heads, head_dim // config.partial_rotary_factor]
+        query_rot, key_rot = self.rotary_emb(query_rot, key_rot, position_ids)
+
+        # [batch_size, seq_length, num_heads, head_dim]
+        query_states = torch.cat((query_rot, query_pass), dim=-1)
+        key_states = torch.cat((key_rot, key_pass), dim=-1)
+        ... # TODO copy the rest of the function? if we do this it's unusable
+
+
+
+PersimmonSdpaAttention = PersimmonConverter.register("PersimmonSdpaAttention", LlamaAttention) 
+PersimmonFlashAttention2 = PersimmonConverter.register("PersimmonFlashAttention2", LlamaAttention) 
+
+COHERE_ATTENTION_CLASSES = {"eager": PersimmonAttention, "flash_attention_2": PersimmonFlashAttention2, "sdpa": PersimmonSdpaAttention}
+
+PersimmonConverter.register("PersimmonDecoderLayer", LlamaDecoderLayer) 
+PersimmonConverter.register("PersimmonPreTrainedModel", LlamaPreTrainedModel)
+
+PersimmonConverter.register("PersimmonModel", LlamaModel)
+PersimmonConverter.register("PersimmonForCausalLM", LlamaForCausalLM)

src/transformers/models/stablelm/diff_stablelm.py

@@ -0,0 +1,395 @@
+from typing import Tuple
+from transformers.models.llama.configuration_llama import LlamaConfig
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from transformers import StableLmConfig
+from transformers.utils import ModelConverter
+
+StableLmConverter = ModelConverter(__file__)
+
+StableLmRMSNorm = StableLmConverter.register("StableLmRMSNorm", LlamaRMSNorm)
+StarcoderRotaryEmbedding = StableLmConverter.register("StarcoderRotaryEmbedding", LlamaRotaryEmbedding)
+StableLmMLP = StableLmConverter.register("StableLmMLP", LlamaMLP)
+
+
+class StableLmLayerNormPerHead(nn.Module):
+    def __init__(self, dim, num_heads, eps=1e-5, bias=False):
+        super().__init__()
+        self.dim = dim
+        self.num_heads = num_heads
+        self.norms = nn.ModuleList([nn.LayerNorm(dim, eps=eps, bias=bias) for _ in range(self.num_heads)])
+
+    def forward(self, hidden_states: torch.Tensor):
+        # Split along the num_heads axis to get per-head inputs
+        # [batch_size, num_heads, seq_len, head_dim] -> [batch_size, 1, seq_len, head_dim] * num_heads
+        states_per_heads = torch.split(hidden_states, 1, dim=1)
+        # Normalize and merge the heads back together
+        return torch.cat([norm(hidden_states) for norm, hidden_states in zip(self.norms, states_per_heads)], dim=1)
+
+class StableLmAttention(LlamaAttention):
+    def __init__(self, config: LlamaConfig, layer_idx: int | None = None):
+        super().__init__(config, layer_idx) # here call to super means
+                                            # we should copy super
+        self.qk_layernorm = config.qk_layernorm
+        self.q_layernorm = StableLmLayerNormPerHead(self.head_dim, self.num_heads, eps=config.layer_norm_eps)
+        self.k_layernorm = StableLmLayerNormPerHead(
+            self.head_dim, self.num_key_value_heads, eps=config.layer_norm_eps
+        )
+        self.attention_dropout = nn.Dropout(config.attention_dropout)
+
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        bsz, q_len, _ = hidden_states.size()
+
+        query_states = self.q_proj(hidden_states)
+        key_states = self.k_proj(hidden_states)
+        value_states = self.v_proj(hidden_states)
+
+        query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
+        key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+        value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+
+        if self.qk_layernorm:
+            query_states = self.q_layernorm(query_states)
+            key_states = self.k_layernorm(key_states)
+
+        kv_seq_len = key_states.shape[-2]
+        if past_key_value is not None:
+            if self.layer_idx is None:
+                raise ValueError(
+                    f"The cache structure has changed since version v4.36. If you are using {self.__class__.__name__} "
+                    "for auto-regressive decoding with k/v caching, please make sure to initialize the attention class "
+                    "with a layer index."
+                )
+            kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
+        cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
+
+        # Partial rotary embedding
+        query_rot, query_pass = (
+            query_states[..., : self.rotary_emb.dim],
+            query_states[..., self.rotary_emb.dim :],
+        )
+        key_rot, key_pass = (
+            key_states[..., : self.rotary_emb.dim],
+            key_states[..., self.rotary_emb.dim :],
+        )
+        # [batch_size, seq_length, num_heads, head_dim // config.partial_rotary_factor]
+        query_rot, key_rot = self.rotary_emb(query_rot, key_rot, cos, sin, position_ids)
+
+        # [batch_size, seq_length, num_heads, head_dim]
+        query_states = torch.cat((query_rot, query_pass), dim=-1)
+        key_states = torch.cat((key_rot, key_pass), dim=-1)
+
+        if past_key_value is not None:
+            # Specific to RoPE models with partial rotation
+            cache_kwargs = {"sin": sin, "cos": cos, "partial_rotation_size": self.rotary_emb.dim}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        # Repeat k/v heads if n_kv_heads < n_heads
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
+
+        if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
+            raise ValueError(
+                f"Attention weights should be of size {(bsz, self.num_heads, q_len, kv_seq_len)}, but is"
+                f" {attn_weights.size()}"
+            )
+
+        if attention_mask is not None:
+            if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
+                raise ValueError(
+                    f"Attention mask should be of size {(bsz, 1, q_len, kv_seq_len)}, but is {attention_mask.size()}"
+                )
+            attn_weights = attn_weights + attention_mask
+
+        # upcast attention to fp32
+        attn_weights = nn.functional.softmax(attn_weights, dtype=torch.float32, dim=-1).to(query_states.dtype)
+        attn_weights = self.attention_dropout(attn_weights)
+
+        attn_output = torch.matmul(attn_weights, value_states)
+
+        if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
+            raise ValueError(
+                f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is"
+                f" {attn_output.size()}"
+            )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+        attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
+
+        attn_output = self.o_proj(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+class StableLmSdpaAttention(StableLmAttention):
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        if output_attentions:
+            # TODO: Improve this warning with e.g. `model.config.attn_implementation = "manual"` once this is implemented.
+            logger.warning_once(
+                "StableLmModel is using StableLmSdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True`. Falling back to the manual attention implementation, "
+                'but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.'
+            )
+            return super().forward(
+                hidden_states=hidden_states,
+                attention_mask=attention_mask,
+                position_ids=position_ids,
+                past_key_value=past_key_value,
+                output_attentions=output_attentions,
+                use_cache=use_cache,
+            )
+
+        bsz, q_len, _ = hidden_states.size()
+
+        query_states = self.q_proj(hidden_states)
+        key_states = self.k_proj(hidden_states)
+        value_states = self.v_proj(hidden_states)
+
+        query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
+        key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+        value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+
+        if self.qk_layernorm:
+            query_states = self.q_layernorm(query_states)
+            key_states = self.k_layernorm(key_states)
+
+        kv_seq_len = key_states.shape[-2]
+        if past_key_value is not None:
+            if self.layer_idx is None:
+                raise ValueError(
+                    f"The cache structure has changed since version v4.36. If you are using {self.__class__.__name__} "
+                    "for auto-regressive decoding with k/v caching, please make sure to initialize the attention class "
+                    "with a layer index."
+                )
+            kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
+        cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
+
+        # Partial rotary embedding
+        query_rot, query_pass = (
+            query_states[..., : self.rotary_emb.dim],
+            query_states[..., self.rotary_emb.dim :],
+        )
+        key_rot, key_pass = (
+            key_states[..., : self.rotary_emb.dim],
+            key_states[..., self.rotary_emb.dim :],
+        )
+        # [batch_size, seq_length, num_heads, head_dim // config.partial_rotary_factor]
+        query_rot, key_rot = apply_rotary_pos_emb(query_rot, key_rot, cos, sin, position_ids)
+
+        # [batch_size, seq_length, num_heads, head_dim]
+        query_states = torch.cat((query_rot, query_pass), dim=-1)
+        key_states = torch.cat((key_rot, key_pass), dim=-1)
+
+        if past_key_value is not None:
+            # Specific to RoPE models with partial rotation
+            cache_kwargs = {"sin": sin, "cos": cos, "partial_rotation_size": self.rotary_emb.dim}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        # Repeat k/v heads if n_kv_heads < n_heads
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        # SDPA with memory-efficient backend is currently (torch==2.1.2) bugged with non-contiguous inputs with custom attn_mask,
+        # Reference: https://github.com/pytorch/pytorch/issues/112577.
+        if query_states.device.type == "cuda" and attention_mask is not None:
+            query_states = query_states.contiguous()
+            key_states = key_states.contiguous()
+            value_states = value_states.contiguous()
+
+        attn_output = torch.nn.functional.scaled_dot_product_attention(
+            query_states,
+            key_states,
+            value_states,
+            attn_mask=attention_mask,
+            dropout_p=self.attention_dropout.p if self.training else 0.0,
+            # The q_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case q_len == 1.
+            is_causal=self.is_causal and attention_mask is None and q_len > 1,
+        )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+        attn_output = attn_output.view(bsz, q_len, self.hidden_size)
+
+        attn_output = self.o_proj(attn_output)
+
+        return attn_output, None, past_key_value
+
+class StableLmFlashAttention2(LlamaFlashAttention2):
+    """
+    StableLM flash attention module. This module inherits from `StableLmAttention` as the weights of the module stays
+    untouched. The only required change would be on the forward pass where it needs to correctly call the public API of
+    flash attention and deal with padding tokens in case the input contains any of them.
+    """
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.LongTensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        output_attentions = False
+
+        bsz, q_len, _ = hidden_states.size()
+
+        query_states = self.q_proj(hidden_states)
+        key_states = self.k_proj(hidden_states)
+        value_states = self.v_proj(hidden_states)
+
+        # Flash attention requires the input to have the shape
+        # batch_size x seq_length x head_dim x hidden_dim
+        # therefore we just need to keep the original shape
+        query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
+        key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+        value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+
+        if self.qk_layernorm:
+            query_states = self.q_layernorm(query_states)
+            key_states = self.k_layernorm(key_states)
+
+        kv_seq_len = key_states.shape[-2]
+        if past_key_value is not None:
+            if self.layer_idx is None:
+                raise ValueError(
+                    f"The cache structure has changed since version v4.36. If you are using {self.__class__.__name__} "
+                    "for auto-regressive decoding with k/v caching, please make sure to initialize the attention class "
+                    "with a layer index."
+                )
+            kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
+        cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
+
+        # Partial rotary embedding
+        query_rot, query_pass = (
+            query_states[..., : self.rotary_emb.dim],
+            query_states[..., self.rotary_emb.dim :],
+        )
+        key_rot, key_pass = (
+            key_states[..., : self.rotary_emb.dim],
+            key_states[..., self.rotary_emb.dim :],
+        )
+        query_rot, key_rot = self.rotary_emb(query_rot, key_rot, cos, sin, position_ids)
+
+        # [batch_size, seq_length, num_heads, head_dim]
+        query_states = torch.cat((query_rot, query_pass), dim=-1)
+        key_states = torch.cat((key_rot, key_pass), dim=-1)
+
+        if past_key_value is not None:
+            cache_kwargs = {"sin": sin, "cos": cos, "partial_rotation_size": self.rotary_emb.dim}
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        # TODO: These transpose are quite inefficient but Flash Attention requires the layout [batch_size, sequence_length, num_heads, head_dim]. We would need to refactor the KV cache
+        # to be able to avoid many of these transpose/reshape/view.
+        query_states = query_states.transpose(1, 2)
+        key_states = key_states.transpose(1, 2)
+        value_states = value_states.transpose(1, 2)
+
+        dropout_rate = self.attention_dropout.p if self.training else 0.0
+
+        attn_output = self._flash_attention_forward(
+            query_states,
+            key_states,
+            value_states,
+            attention_mask,
+            q_len,
+            dropout=dropout_rate,
+        )
+
+        attn_output = attn_output.reshape(bsz, q_len, self.hidden_size).contiguous()
+        attn_output = self.o_proj(attn_output)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+
+StableLm_ATTENTION_CLASSES = {"eager": StableLmAttention, "flash_attention_2": StableLmFlashAttention2, "sdpa": StableLmSdpaAttention}
+
+class StableLmDecoderLayer(nn.Module):
+    def __init__(self, config: StableLmConfig, layer_idx: int):
+        super().__init__()
+        self.use_parallel_residual = config.use_parallel_residual
+        self.hidden_size = config.hidden_size
+        self.self_attn = StableLm_ATTENTION_CLASSES[config._attn_implementation](config, layer_idx=layer_idx)
+        self.mlp = StableLmMLP(config)
+        self.input_layernorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.post_attention_layernorm = None
+        if not self.use_parallel_residual:
+            self.post_attention_layernorm = nn.LayerNorm(config.hidden_size, eps=config.layer_norm_eps)
+        self.dropout = nn.Dropout(config.hidden_dropout)
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Tuple[torch.Tensor]] = None,
+        output_attentions: Optional[bool] = False,
+        use_cache: Optional[bool] = False,
+    ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]:
+        residual = hidden_states
+
+        hidden_states = self.input_layernorm(hidden_states)
+
+        # Self Attention
+        self_attn_output, self_attn_weights, present_key_value = self.self_attn(
+            hidden_states=hidden_states,
+            attention_mask=attention_mask,
+            position_ids=position_ids,
+            past_key_value=past_key_value,
+            output_attentions=output_attentions,
+            use_cache=use_cache,
+        )
+
+        # copied from transformers.models.gpt_neox.modeling_gpt_neox.GPTNeoXLayer.forward
+        if self.use_parallel_residual:
+            # x = x + attn(ln1(x)) + mlp(ln1(x))
+            # Fully Connected
+            mlp_output = self.mlp(hidden_states)
+            mlp_output = self.dropout(mlp_output)
+            hidden_states = residual + self_attn_output + mlp_output
+        else:
+            # x = x + attn(ln1(x))
+            # x = x + mlp(ln2(x))
+            residual = residual + self_attn_output
+            # Fully Connected
+            mlp_output = self.mlp(self.post_attention_layernorm(residual))
+            mlp_output = self.dropout(mlp_output)
+            hidden_states = residual + mlp_output
+
+        outputs = (hidden_states,)
+
+        if output_attentions:
+            outputs += (self_attn_weights,)
+
+        if use_cache:
+            outputs += (present_key_value,)
+
+        return outputs
+
+StableLmPreTrainedModel = StableLmConverter.register("StableLmPreTrainedModel", LlamaPreTrainedModel)
+StableLmdModel = StableLmConverter.register("StableLmdModel", LlamaModel)
+StableLmForCausalLM = StableLmConverter.register("StableLmForCausalLM", LlamaForCausalLM)
+StableLmForSequenceClassification = StableLmConverter.register("StableLmForSequenceClassification", LlamaForSequenceClassification)

src/transformers/models/starcoder2/diff_starcoder2.py

@@ -0,0 +1,166 @@
+# coding=utf-8
+# Copyright 2024 Google Inc. 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.
+from typing import List, Tuple
+from torch import FloatTensor, LongTensor, Tensor
+from transformers.modeling_outputs import BaseModelOutputWithPast
+from transformers.models.llama.configuration_llama import LlamaConfig
+from transformers.models.llama.modeling_llama import *
+import torch.nn as nn
+from transformers import Starcoder2Config
+from transformers.utils import ModelConverter
+
+Starcoder2Converter = ModelConverter(__file__)
+
+Starcoder2RMSNorm = Starcoder2Converter.register("Starcoder2RMSNorm", LlamaRMSNorm)
+StarcoderRotaryEmbedding = Starcoder2Converter.register("StarcoderRotaryEmbedding", LlamaRotaryEmbedding)
+
+class Starcoder2MLP(nn.Module):
+    def __init__(self, config: Starcoder2Config):
+        super().__init__()
+        embed_dim = config.hidden_size
+        self.c_fc = nn.Linear(embed_dim, config.intermediate_size, bias=config.use_bias)
+        self.c_proj = nn.Linear(config.intermediate_size, embed_dim, bias=config.use_bias)
+        self.act = ACT2FN[config.hidden_act]
+        self.residual_dropout = config.residual_dropout
+
+    def forward(self, hidden_states: Optional[Tuple[torch.FloatTensor]]) -> torch.FloatTensor:
+        hidden_states = self.c_fc(hidden_states)
+        hidden_states = self.act(hidden_states)
+        hidden_states = self.c_proj(hidden_states)
+        hidden_states = nn.functional.dropout(hidden_states, p=self.residual_dropout, training=self.training)
+        return hidden_states
+
+# TODO either we support this, or we don't allow call to super?
+# if part of the super is used, then we are fucked. Let's restrict this to init?
+
+# TODO if a class is not registered, the original should be copied with replaces?
+# Copied form where? No.
+# But then how do we check the architecture etc.
+
+# TODO do we support multiple inheritance? 
+# This will depend on whether we usually copy from more than one module
+# Mixtral for example? 
+
+class Starcoder2Attention(LlamaAttention):
+    def __init__(self, config: LlamaConfig, layer_idx: int | None = None):
+        super().__init__(config, layer_idx) # here call to super means
+        self.attention_dropout = config.attention_dropout
+
+    def forward(
+        self,
+        hidden_states: torch.Tensor,
+        attention_mask: Optional[torch.Tensor] = None,
+        position_ids: Optional[torch.LongTensor] = None,
+        past_key_value: Optional[Cache] = None,
+        output_attentions: bool = False,
+        use_cache: bool = False,
+        **kwargs,
+    ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
+        if "padding_mask" in kwargs:
+            warnings.warn(
+                "Passing `padding_mask` is deprecated and will be removed in v4.37. Please make sure use `attention_mask` instead.`"
+            )
+        bsz, q_len, _ = hidden_states.size()
+
+        query_states = self.q_proj(hidden_states)
+        key_states = self.k_proj(hidden_states)
+        value_states = self.v_proj(hidden_states)
+
+        query_states = query_states.view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
+        key_states = key_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+        value_states = value_states.view(bsz, q_len, self.num_key_value_heads, self.head_dim).transpose(1, 2)
+
+        kv_seq_len = key_states.shape[-2]
+        if past_key_value is not None:
+            if self.layer_idx is None:
+                raise ValueError(
+                    f"The cache structure has changed since version v4.36. If you are using {self.__class__.__name__} "
+                    "for auto-regressive decoding with k/v caching, please make sure to initialize the attention class "
+                    "with a layer index."
+                )
+            kv_seq_len += past_key_value.get_usable_length(kv_seq_len, self.layer_idx)
+        cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
+        query_states, key_states = self.rotary_emb(query_states, key_states, cos, sin, position_ids)
+
+        if past_key_value is not None:
+            cache_kwargs = {"sin": sin, "cos": cos}  # Specific to RoPE models
+            key_states, value_states = past_key_value.update(key_states, value_states, self.layer_idx, cache_kwargs)
+
+        # repeat k/v heads if n_kv_heads < n_heads
+        key_states = repeat_kv(key_states, self.num_key_value_groups)
+        value_states = repeat_kv(value_states, self.num_key_value_groups)
+
+        attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
+
+        if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
+            raise ValueError(
+                f"Attention weights should be of size {(bsz, self.num_heads, q_len, kv_seq_len)}, but is"
+                f" {attn_weights.size()}"
+            )
+
+        if attention_mask is not None:
+            if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
+                raise ValueError(
+                    f"Attention mask should be of size {(bsz, 1, q_len, kv_seq_len)}, but is {attention_mask.size()}"
+                )
+
+            attn_weights = attn_weights + attention_mask
+
+        # upcast attention to fp32
+        attn_weights = nn.functional.softmax(attn_weights, dim=-1, dtype=torch.float32).to(query_states.dtype)
+        attn_weights = nn.functional.dropout(attn_weights, p=self.attention_dropout, training=self.training)
+        attn_output = torch.matmul(attn_weights, value_states)
+
+        if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
+            raise ValueError(
+                f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is"
+                f" {attn_output.size()}"
+            )
+
+        attn_output = attn_output.transpose(1, 2).contiguous()
+        attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
+
+        attn_output = self.o_proj(attn_output)
+        attn_output = nn.functional.dropout(attn_output, p=self.residual_dropout, training=self.training)
+
+        if not output_attentions:
+            attn_weights = None
+
+        return attn_output, attn_weights, past_key_value
+
+Starcoder2SdpaAttention = Starcoder2Converter.register("Starcoder2SdpaAttention", LlamaAttention) 
+Starcoder2FlashAttention2 = Starcoder2Converter.register("Starcoder2FlashAttention2", LlamaAttention) 
+
+STARCODER2_ATTENTION_CLASSES = {"eager": Starcoder2Attention, "flash_attention_2": Starcoder2FlashAttention2, "sdpa": Starcoder2SdpaAttention}
+
+
+Starcoder2DecoderLayer = Starcoder2Converter.register("Starcoder2DecoderLayer", LlamaDecoderLayer) 
+Starcoder2PreTrainedModel = Starcoder2Converter.register("Starcoder2PreTrainedModel", LlamaPreTrainedModel)
+
+class Starcoder2Model(LlamaModel):
+    def __init__(self, config):
+        super().__init__(config)
+        self.embedding_dropout = config.embedding_dropout
+
+    def forward(self, input_ids: LongTensor = None, attention_mask: Tensor | None = None, position_ids: LongTensor | None = None, past_key_values: List[FloatTensor] | None = None, inputs_embeds: FloatTensor | None = None, use_cache: bool | None = None, output_attentions: bool | None = None, output_hidden_states: bool | None = None, return_dict: bool | None = None, cache_position: LongTensor | None = None) -> Tuple | BaseModelOutputWithPast:
+        if inputs_embeds is None: 
+            inputs_embeds = self.embed_tokens(input_ids)
+        hidden_states = inputs_embeds
+        hidden_states = nn.functional.dropout(hidden_states, p=self.embedding_dropout, training=self.training)
+        return super().forward(None, attention_mask, position_ids, past_key_values, inputs_embeds, use_cache, output_attentions, output_hidden_states, return_dict, cache_position)
+
+Starcoder2ForCausalLM = Starcoder2Converter.register("Starcoder2ForCausalLM", LlamaForCausalLM)
+Starcoder2ForSequenceClassification = Starcoder2Converter.register("Starcoder2ForSequenceClassification", LlamaForSequenceClassification)

src/transformers/utils/__init__.py

@@ -215,7 +215,7 @@ from .peft_utils import (
     check_peft_version,
     find_adapter_config_file,
 )
-
+from .model_converter import ModelConverter
 
 WEIGHTS_NAME = "pytorch_model.bin"
 WEIGHTS_INDEX_NAME = "pytorch_model.bin.index.json"

src/transformers/utils/model_converter.py

@@ -0,0 +1,57 @@
+"""
+running this script on `src/transformers/models/**_diff.py` should produce the equivalent single model single files
+1. Iterate though `**_diff.py` files
+2. How to handle the imports?
+    a. `model_type` should always be present?
+    b. `ConfigClass` should be defined as well?
+3. Copy each class and function one by one.
+    a. if there is a class registered for this file like `@__file__.register(MyNewClass, OldClass)`
+    then copy the content of `OldClass`, replacing all names of `Old` with `MyNew`.
+    Also copy the decorators that are on top of this class.
+    b. if there is inheritance, copy non-overloaded functions from base, and overloaded from non base.
+4. Register things?
+new = type("new_class", (torch.nn.Linear,),{})
+new__main__.new_class
+new(10,10)
+new_class(in_features=10, out_features=10, bias=True)
+CohereConverter = ModelConverter(__file__)
+CohereMLP = CohereConverter.register("CohereMLP", LlamaMLP)
+CohereMLP
+<class 'transformers.models.cohere.modeling_cohere.CohereMLP'>
+CohereMLP(LlamaConfig())
+CohereMLP(
+  (gate_proj): Linear(in_features=4096, out_features=11008, bias=False)
+  (up_proj): Linear(in_features=4096, out_features=11008, bias=False)
+  (down_proj): Linear(in_features=11008, out_features=4096, bias=False)
+  (act_fn): SiLU()
+)
+>>> CohereMLP(LlamaConfig())(torch.ones(1,1,4096))
+How to deal with submodules?
+CohereSdpaAttention(
+  (q_proj): Linear(in_features=4096, out_features=4096, bias=False)
+  (k_proj): Linear(in_features=4096, out_features=4096, bias=False)
+  (v_proj): Linear(in_features=4096, out_features=4096, bias=False)
+  (o_proj): Linear(in_features=4096, out_features=4096, bias=False)
+  (rotary_emb): LlamaRotaryEmbedding()
+)
+"""
+import regex as re
+class ModelConverter:
+
+    def __init__(self, file):
+        self.diff_file = file
+        self.model_name = re.search(r'models/(.*?)/diff', self.diff_file).group(1)
+        self.modeling_file = file.replace("diff", "modeling")
+        self.registered_classes = {}
+        self.modules_to_import = []
+    def register(self, new_class, old_class):
+        # registering. Returns the old class to be usable with a new name
+        self.registered_classes[new_class] = old_class
+        self.modules_to_import.append([old_class, old_class.__module__])
+        new_class = type(new_class, (old_class,), {})
+        base_model_name = re.search(r'models\.(.*?)\.modeling', old_class.__module__).group(1)
+        new_class.__module__ = re.sub(base_model_name, self.model_name, old_class.__module__)
+        return new_class
+
+    def __repr__(self) -> str:
+        return f"ModelConverter({self.diff_file}, {self.model_name}, {self.registered_classes})"

utils/convert_diff_model.py

@@ -0,0 +1,148 @@
+import importlib
+import argparse
+import glob
+from transformers import MODEL_NAMES_MAPPING
+import regex as re
+import inspect
+# pattern = re.compile(r'(class|def|XXXConverter\.register)\s+[\w.()]+\s*:(\s*(?:[^class|def|XXXConverter\.register]|\n)+)', re.MULTILINE)
+# For each and every diff files we should import all packages from the modules that are imported.
+# pattern = r"((    [\s\S]*?)\n\n(?=    \S))|((    [\s\S]*?)(?=\Z))" is super important
+
+# TODO in order to get everything from LLAMA we need to copy each line from Llama
+# only updating the attention classes. 
+# Need to keep the order correct
+# TODO the imports here are not correctly done. We should dynamically import what we need
+APACHE_LICENCE = """# 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.
+"""
+# 1. all the imports from the original file should be copied until end of header? __HEADER__
+# with open(CohereConverter.original_file, 'r') as file, open("result.py", "w+") as modeling:
+#         pass
+# TODO also copy and import from all modules in CohereConverter.modules_to_import to be able to use inspect
+def replace_super_calls_in_method(method_body, parent_method_body, method_name):
+    # Indent parent method body to match the child's method indentation
+    indent = re.match(r'(\s*)def ', method_body).group(1)
+    indented_parent_method_body = "\n".join([indent + line if line.strip() else line for line in parent_method_body.split('\n')])
+    method_name = method_name.strip()
+    # Handle super().method_name(args) and return super().method_name(args)
+    super_call_pattern = re.compile(r'(\s*)return super\(\)\.' + method_name + r'\((.*?)\)')
+    method_body = super_call_pattern.sub(r'\1return (\2\n' + indented_parent_method_body + r'\1)', method_body)
+
+    super_call_pattern_no_return = re.compile(r'(\s*)super\(\)\.' + method_name + r'\((.*?)\)')
+    method_body = super_call_pattern_no_return.sub(r'\1\2\n' + indented_parent_method_body, method_body)
+
+    return method_body
+# 2. Write all the classes. Use the `CohereConverter` class for this.
+
+
+def create_single_model_file(converter):
+    if hasattr(converter, "diff_file"):
+        model_identifier = converter.diff_file.split("diff_")
+        # temporarily add the source to the path in order to load everything?
+        # 1. Import all modules from the registered classes
+        modules = set([ _class.__module__ for _class in converter.registered_classes.values()]) or set()
+        for module in modules | {re.sub(r'.*src/(.*)\.py', r'\1', converter.diff_file).replace('/', '.')}:
+            modeling_ = importlib.import_module(module)
+            globals().update({k: getattr(modeling_, k) for k in modeling_.__dict__.keys()})
+
+        with open(converter.diff_file, 'r') as file, open(f"{model_identifier[0]}modeling_{model_identifier[1]}", "w+") as modeling:
+            modeling.write(APACHE_LICENCE)
+            function_set = {}
+            for line in file:
+                    if "Converter.register" in line: # TODO use map() to map lines to this
+                        # write the code of the original model
+                        class_to_use, old_class = re.search(r'Converter\.register\(\"(.*?)\", (.*?)\)', line).groups()
+                        model_identifier_camel = re.findall(r'[A-Z][a-z0-9]*', class_to_use)[0]
+                        old_model_identifier_camel = re.findall(r'[A-Z][a-z0-9]*', old_class)[0]
+                        # import all necessary modules from the path:
+
+                        source_code = inspect.getsource(converter.registered_classes[class_to_use]).replace(old_class, class_to_use)
+                        source_code = source_code.replace(old_model_identifier_camel, model_identifier_camel)
+                        modeling.write(source_code)
+                        modeling.write("\n")
+
+                    elif match:=re.match(r"class (\w+)\((\w+)\):", line):
+                        class_name, parent_class = match.groups()
+                        pattern = re.compile( r"(\ {4}(?:[\S\s\ \n]*?)(?=\n\ ^[\) ]|\n\n    (?:def|@)|\Z))", re.MULTILINE)
+                        parent_class_def = inspect.getsource(eval(parent_class))
+                        modeling.write(parent_class_def.split('\n')[0].replace(parent_class,class_name)+"\n")
+
+                        function_name_pattern = r"(?=    def ([\S]*)\()"
+                        function_body_pattern = r"(\ {4}(?:[\S\s\ \n]*?)(?=\n\ ^[\) ]|\n\n    (?:def|@)|\Z))"
+
+                        pattern = re.compile(function_body_pattern)
+                        matches = pattern.finditer(parent_class_def)
+                        parent_function_set = {}
+                        for match in matches:
+                            full_function = match.group()
+                            if "def" in full_function:
+                                parent_function_set[full_function.split("def")[1].split("(")[0]] = full_function
+                            else:
+                                parent_function_set[full_function] = full_function
+
+                        parent_identifier_camel = re.findall(r'[A-Z][a-z0-9]*', parent_class)[0]
+                        child_identifier_camel = re.findall(r'[A-Z][a-z0-9]*', class_name)[0]
+                        print(f"`{class_name}` -> `{parent_class}`")
+
+                        child_function_set = parent_function_set.copy()
+                        class_def = inspect.getsource(eval(class_name))
+                        matches = pattern.finditer(class_def)
+                        for match in matches:
+                            # TODO handle call to super!
+                            full_function = match.group()
+                            if "def" in full_function:
+                                function_name = full_function.split("def")[1].split("(")[0]
+
+                                if (f"super()." in full_function or f"return super()." in full_function) and parent_identifier_camel != child_identifier_camel:
+                                    replaced_function = replace_super_calls_in_method(full_function,
+                                                                                      parent_function_set.get(function_name,
+                                                                                                              ""),
+                                                                                      function_name)
+                                    child_function_set[function_name] = replaced_function
+                                else:
+                                    child_function_set[function_name] = full_function
+                            else:
+                                child_function_set[full_function] = full_function
+
+                        modeling.write("\n".join(child_function_set.values())) # TODO we wrote the code, next lines shall be ignored
+                        modeling.write("\n")
+
+                    elif "= ModelConverter(__file__)" in line:
+                        pass # don't write the converter to the result file
+                    elif line not in "".join(function_set.values()) or line=="\n":
+                        modeling.write(line)
+
+
+def dynamically_import_object(module_path, object_name):
+    try:
+        module = importlib.import_module(module_path)
+        obj = getattr(module, object_name)
+        return obj
+    except (ImportError, AttributeError) as e:
+        print(f"Failed to import object '{object_name}' from module '{module_path}'")
+        print(e)
+
+
+# 3. Apply ruff fix to remove unused imports
+# 4. Run a tiny test to import from this new file.
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--files_to_parse", default="all", help="A list of `diff_xxxx` files that should be converted to single model file")
+    args = parser.parse_args()
+    if args.files_to_parse == "all":
+        args.files_to_parse = glob.glob("src/transformers/models/**/diff_*.py", recursive=True)
+    for file_name in args.files_to_parse:
+        print(f"Converting {file_name} to a single model single file format")
+        module_path = file_name.replace("/",".").replace(".py","").replace("src.","")
+        model_name = MODEL_NAMES_MAPPING[module_path.split('_')[-1]]
+        converter = dynamically_import_object(module_path, f"{model_name}Converter")
+        create_single_model_file(converter)