cArlIcon commited on
Commit
d729eec
1 Parent(s): 40ad1b8

add Yi-6B model and tokenizer

Browse files
config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "YiForCausalLM"
4
+ ],
5
+ "auto_map": {
6
+ "AutoConfig": "configuration_yi.YiConfig",
7
+ "AutoModel": "modeling_yi.YiModel",
8
+ "AutoModelForCausalLM":"modeling_yi.YiForCausalLM"
9
+ },
10
+ "bos_token_id": 1,
11
+ "eos_token_id": 2,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 11008,
16
+ "max_position_embeddings": 200000,
17
+ "model_type": "Yi",
18
+ "num_attention_heads": 32,
19
+ "num_hidden_layers": 32,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 0,
22
+ "rms_norm_eps": 1e-05,
23
+ "rope_theta": 5000000.0,
24
+ "tie_word_embeddings": false,
25
+ "torch_dtype": "bfloat16",
26
+ "transformers_version": "4.34.0",
27
+ "use_cache": true,
28
+ "vocab_size": 64000
29
+ }
configuration_yi.py ADDED
@@ -0,0 +1,121 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """ Yi model configuration"""
2
+ from transformers.configuration_utils import PretrainedConfig
3
+ from transformers.utils import logging
4
+
5
+ logger = logging.get_logger(__name__)
6
+
7
+ Yi_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
8
+
9
+
10
+ class YiConfig(PretrainedConfig):
11
+ r"""
12
+ This is the configuration class to store the configuration of a [`YiModel`]. It is used to instantiate an Yi
13
+ model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
14
+ defaults will yield a similar configuration to that of the Yi model.
15
+
16
+ Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
17
+ documentation from [`PretrainedConfig`] for more information.
18
+
19
+
20
+ Args:
21
+ vocab_size (`int`, *optional*, defaults to 64000):
22
+ Vocabulary size of the Yi model. Defines the number of different tokens that can be represented by the
23
+ `inputs_ids` passed when calling [`YiModel`]
24
+ hidden_size (`int`, *optional*, defaults to 4096):
25
+ Dimension of the hidden representations.
26
+ intermediate_size (`int`, *optional*, defaults to 11008):
27
+ Dimension of the MLP representations.
28
+ num_hidden_layers (`int`, *optional*, defaults to 32):
29
+ Number of hidden layers in the Transformer encoder.
30
+ num_attention_heads (`int`, *optional*, defaults to 32):
31
+ Number of attention heads for each attention layer in the Transformer encoder.
32
+ num_key_value_heads (`int`, *optional*):
33
+ This is the number of key_value heads that should be used to implement Grouped Query Attention. If
34
+ `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
35
+ `num_key_value_heads=1 the model will use Multi Query Attention (MQA) otherwise GQA is used. When
36
+ converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
37
+ by meanpooling all the original heads within that group. For more details checkout [this
38
+ paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to
39
+ `num_attention_heads`.
40
+ hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
41
+ The non-linear activation function (function or string) in the decoder.
42
+ max_position_embeddings (`int`, *optional*, defaults to 4096):
43
+ The maximum sequence length that this model might ever be used with. Typically set this to something large
44
+ just in case (e.g., 512 or 1024 or 2048 or 4096).
45
+ initializer_range (`float`, *optional*, defaults to 0.02):
46
+ The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
47
+ rms_norm_eps (`float`, *optional*, defaults to 1e-5):
48
+ The epsilon used by the rms normalization layers.
49
+ use_cache (`bool`, *optional*, defaults to `True`):
50
+ Whether or not the model should return the last key/values attentions (not used by all models). Only
51
+ relevant if `config.is_decoder=True`.
52
+ tie_word_embeddings(`bool`, *optional*, defaults to `False`):
53
+ Whether to tie weight embeddings
54
+ output_attentions (`bool`, *optional*, defaults to `False`):
55
+ Whether or not to output attentions.
56
+ rope_theta (`float`, *optional*, defaults to 5000000.0):
57
+ The base period of the RoPE embeddings.
58
+ Example:
59
+
60
+ ```python
61
+ >>> from transformers import YiModel, YiConfig
62
+
63
+ >>> # Initializing a Yi style configuration
64
+ >>> configuration = YiConfig()
65
+
66
+ >>> # Initializing a model from the Yi style configuration
67
+ >>> model = YiModel(configuration)
68
+
69
+ >>> # Accessing the model configuration
70
+ >>> configuration = model.config
71
+ ```"""
72
+ model_type = "Yi"
73
+ keys_to_ignore_at_inference = ["past_key_values"]
74
+
75
+ def __init__(
76
+ self,
77
+ vocab_size=64000,
78
+ hidden_size=4096,
79
+ intermediate_size=11008,
80
+ num_hidden_layers=32,
81
+ num_attention_heads=32,
82
+ num_key_value_heads=4,
83
+ hidden_act="silu",
84
+ max_position_embeddings=4096,
85
+ initializer_range=0.02,
86
+ rms_norm_eps=1e-5,
87
+ use_cache=True,
88
+ pad_token_id=0,
89
+ bos_token_id=1,
90
+ eos_token_id=2,
91
+ tie_word_embeddings=False,
92
+ output_attentions=False,
93
+ rope_theta=5000000.0,
94
+ **kwargs,
95
+ ):
96
+ self.vocab_size = vocab_size
97
+ self.max_position_embeddings = max_position_embeddings
98
+ self.hidden_size = hidden_size
99
+ self.intermediate_size = intermediate_size
100
+ self.num_hidden_layers = num_hidden_layers
101
+ self.num_attention_heads = num_attention_heads
102
+
103
+ # for backward compatibility
104
+ if num_key_value_heads is None:
105
+ num_key_value_heads = num_attention_heads
106
+
107
+ self.num_key_value_heads = num_key_value_heads
108
+ self.hidden_act = hidden_act
109
+ self.initializer_range = initializer_range
110
+ self.rms_norm_eps = rms_norm_eps
111
+ self.use_cache = use_cache
112
+ self.output_attentions = output_attentions
113
+ self.rope_theta = rope_theta
114
+
115
+ super().__init__(
116
+ pad_token_id=pad_token_id,
117
+ bos_token_id=bos_token_id,
118
+ eos_token_id=eos_token_id,
119
+ tie_word_embeddings=tie_word_embeddings,
120
+ **kwargs,
121
+ )
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.34.0"
7
+ }
md5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ e11decc690391b47f62e217a3faca830 pytorch_model-00001-of-00002.bin
2
+ ea6e2b5eddc1416a101361efb286d79c pytorch_model-00002-of-00002.bin
3
+ 291724ef50f729e45d68f474a7755bbc tokenizer.model
modeling_yi.py ADDED
@@ -0,0 +1,1030 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """ PyTorch Yi model."""
2
+ import math
3
+ from typing import List, Optional, Tuple, Union
4
+
5
+ import torch.utils.checkpoint
6
+ from einops import repeat
7
+ from torch import nn
8
+ from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
9
+
10
+ from transformers.activations import ACT2FN
11
+ from transformers.modeling_outputs import (
12
+ BaseModelOutputWithPast,
13
+ CausalLMOutputWithPast,
14
+ SequenceClassifierOutputWithPast,
15
+ )
16
+ from transformers.modeling_utils import PreTrainedModel
17
+ from transformers.pytorch_utils import ALL_LAYERNORM_LAYERS
18
+ from transformers.utils import (
19
+ add_start_docstrings,
20
+ add_start_docstrings_to_model_forward,
21
+ is_flash_attn_available,
22
+ logging,
23
+ replace_return_docstrings,
24
+ )
25
+
26
+ from .configuration_yi import YiConfig
27
+
28
+
29
+ if is_flash_attn_available():
30
+ from flash_attn import flash_attn_func
31
+
32
+
33
+ logger = logging.get_logger(__name__)
34
+
35
+ _CONFIG_FOR_DOC = "YiConfig"
36
+
37
+
38
+ # Copied from transformers.models.bart.modeling_bart._make_causal_mask
39
+ def _make_causal_mask(
40
+ input_ids_shape: torch.Size,
41
+ dtype: torch.dtype,
42
+ device: torch.device,
43
+ past_key_values_length: int = 0,
44
+ ):
45
+ """
46
+ Make causal mask used for bi-directional self-attention.
47
+ """
48
+ bsz, tgt_len = input_ids_shape
49
+ mask = torch.full(
50
+ (tgt_len, tgt_len),
51
+ torch.tensor(torch.finfo(dtype).min, device=device),
52
+ device=device,
53
+ )
54
+ mask_cond = torch.arange(mask.size(-1), device=device)
55
+ mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
56
+ mask = mask.to(dtype)
57
+
58
+ if past_key_values_length > 0:
59
+ mask = torch.cat(
60
+ [
61
+ torch.zeros(
62
+ tgt_len, past_key_values_length, dtype=dtype, device=device
63
+ ),
64
+ mask,
65
+ ],
66
+ dim=-1,
67
+ )
68
+ return mask[None, None, :, :].expand(
69
+ bsz, 1, tgt_len, tgt_len + past_key_values_length
70
+ )
71
+
72
+
73
+ # Copied from transformers.models.bart.modeling_bart._expand_mask
74
+ def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int] = None):
75
+ """
76
+ Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.
77
+ """
78
+ bsz, src_len = mask.size()
79
+ tgt_len = tgt_len if tgt_len is not None else src_len
80
+
81
+ expanded_mask = mask[:, None, None, :].expand(bsz, 1, tgt_len, src_len).to(dtype)
82
+
83
+ inverted_mask = 1.0 - expanded_mask
84
+
85
+ return inverted_mask.masked_fill(
86
+ inverted_mask.to(torch.bool), torch.finfo(dtype).min
87
+ )
88
+
89
+
90
+ class YiRMSNorm(nn.Module):
91
+ def __init__(self, hidden_size, eps=1e-5):
92
+ """
93
+ YiRMSNorm is equivalent to T5LayerNorm
94
+ """
95
+ super().__init__()
96
+ self.weight = nn.Parameter(torch.ones(hidden_size))
97
+ self.variance_epsilon = eps
98
+
99
+ def forward(self, hidden_states):
100
+ input_dtype = hidden_states.dtype
101
+ hidden_states = hidden_states.to(torch.float32)
102
+ variance = hidden_states.pow(2).mean(-1, keepdim=True)
103
+ hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)
104
+
105
+ return self.weight * hidden_states.to(input_dtype)
106
+
107
+
108
+ ALL_LAYERNORM_LAYERS.append(YiRMSNorm)
109
+
110
+
111
+ class YiRotaryEmbedding(torch.nn.Module):
112
+ def __init__(self, dim, max_position_embeddings=4096, base=5000000, device=None):
113
+ super().__init__()
114
+
115
+ self.dim = dim
116
+ self.max_position_embeddings = max_position_embeddings
117
+ self.base = base
118
+
119
+ # Build here to make `torch.jit.trace` work.
120
+ self._set_cos_sin_cache(seq_len=max_position_embeddings, device=device)
121
+
122
+ def _set_cos_sin_cache(self, seq_len, device):
123
+ self.max_seq_len_cached = seq_len
124
+ inv_freq = 1.0 / (
125
+ self.base ** (torch.arange(0, self.dim, 2).float().to(device) / self.dim)
126
+ )
127
+ t = torch.arange(self.max_seq_len_cached, device=device, dtype=torch.float32)
128
+ freqs = torch.einsum("i,j->ij", t, inv_freq)
129
+ # Different from paper, but it uses a different permutation in order to obtain the same calculation
130
+ emb = torch.cat((freqs, freqs), dim=-1)
131
+ self.register_buffer(
132
+ "cos_cached", emb.cos()[None, None, :, :], persistent=False
133
+ )
134
+ self.register_buffer(
135
+ "sin_cached", emb.sin()[None, None, :, :], persistent=False
136
+ )
137
+
138
+ def forward(self, x, seq_len=None):
139
+ # x: [bs, num_attention_heads, seq_len, head_size]
140
+ if seq_len > self.max_seq_len_cached:
141
+ self._set_cos_sin_cache(seq_len=seq_len, device=x.device)
142
+
143
+ return (
144
+ self.cos_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
145
+ self.sin_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
146
+ )
147
+
148
+
149
+ def rotate_half(x):
150
+ """Rotates half the hidden dims of the input."""
151
+ x1 = x[..., : x.shape[-1] // 2]
152
+ x2 = x[..., x.shape[-1] // 2 :]
153
+ return torch.cat((-x2, x1), dim=-1)
154
+
155
+
156
+ def apply_rotary_pos_emb(q, k, cos, sin, position_ids, flash_attn_available):
157
+ # The first two dimensions of cos and sin are always 1, so we can `squeeze` them.
158
+ cos = cos.squeeze(1).squeeze(0) # [seq_len, dim]
159
+ sin = sin.squeeze(1).squeeze(0) # [seq_len, dim]
160
+ expand_dim = 2 if flash_attn_available else 1
161
+ cos = cos[position_ids].unsqueeze(expand_dim) # [bs, seq_len, dim]
162
+ sin = sin[position_ids].unsqueeze(expand_dim) # [bs, seq_len, dim]
163
+ q_embed = (q * cos) + (rotate_half(q) * sin)
164
+ k_embed = (k * cos) + (rotate_half(k) * sin)
165
+ return q_embed, k_embed
166
+
167
+
168
+ class YiMLP(nn.Module):
169
+ def __init__(self, hidden_size: int, intermediate_size: int, hidden_act: str):
170
+ super().__init__()
171
+ self.gate_proj = nn.Linear(hidden_size, intermediate_size, bias=False)
172
+ self.down_proj = nn.Linear(intermediate_size, hidden_size, bias=False)
173
+ self.up_proj = nn.Linear(hidden_size, intermediate_size, bias=False)
174
+ self.act_fn = ACT2FN[hidden_act]
175
+
176
+ def forward(self, x):
177
+ return self.down_proj(self.act_fn(self.gate_proj(x)) * self.up_proj(x))
178
+
179
+
180
+ class YiAttention(nn.Module):
181
+ """Multi-headed attention from 'Attention Is All You Need' paper"""
182
+
183
+ def __init__(self, config: YiConfig):
184
+ super().__init__()
185
+ self.config = config
186
+ self.hidden_size = config.hidden_size
187
+ self.num_heads = config.num_attention_heads
188
+ self.head_dim = self.hidden_size // self.num_heads
189
+ self.num_key_value_heads = config.num_key_value_heads
190
+ self.num_key_value_groups = self.num_heads // self.num_key_value_heads
191
+ self.max_position_embeddings = config.max_position_embeddings
192
+
193
+ if (self.head_dim * self.num_heads) != self.hidden_size:
194
+ raise ValueError(
195
+ f"hidden_size must be divisible by num_heads (got `hidden_size`: {self.hidden_size}"
196
+ f" and `num_heads`: {self.num_heads})."
197
+ )
198
+ self.q_proj = nn.Linear(
199
+ self.hidden_size, self.num_heads * self.head_dim, bias=False
200
+ )
201
+ self.k_proj = nn.Linear(
202
+ self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False
203
+ )
204
+ self.v_proj = nn.Linear(
205
+ self.hidden_size, self.num_key_value_heads * self.head_dim, bias=False
206
+ )
207
+ self.o_proj = nn.Linear(
208
+ self.num_heads * self.head_dim, self.hidden_size, bias=False
209
+ )
210
+
211
+ self.rotary_emb = YiRotaryEmbedding(
212
+ self.head_dim,
213
+ max_position_embeddings=self.max_position_embeddings,
214
+ base=self.config.rope_theta,
215
+ )
216
+
217
+ def forward(
218
+ self,
219
+ hidden_states: torch.Tensor,
220
+ attention_mask: Optional[torch.Tensor] = None,
221
+ position_ids: Optional[torch.LongTensor] = None,
222
+ past_key_value: Optional[Tuple[torch.Tensor]] = None,
223
+ output_attentions: bool = False,
224
+ use_cache: bool = False,
225
+ ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
226
+ bsz, q_len, _ = hidden_states.size()
227
+ flash_attn_available = is_flash_attn_available()
228
+
229
+ query_states = self.q_proj(hidden_states).view(
230
+ bsz, q_len, self.num_heads, self.head_dim
231
+ )
232
+
233
+ key_states = self.k_proj(hidden_states).view(
234
+ bsz, q_len, self.num_key_value_heads, self.head_dim
235
+ )
236
+ value_states = self.v_proj(hidden_states).view(
237
+ bsz, q_len, self.num_key_value_heads, self.head_dim
238
+ )
239
+
240
+ if not flash_attn_available:
241
+ if self.num_key_value_groups > 1:
242
+ key_states = repeat(
243
+ key_states, f"b n h d -> b n (h {self.num_key_value_groups}) d"
244
+ )
245
+ value_states = repeat(
246
+ value_states, f"b n h d -> b n (h {self.num_key_value_groups}) d"
247
+ )
248
+
249
+ # b n h d -> b h n d
250
+ query_states = query_states.transpose(1, 2)
251
+ key_states = key_states.transpose(1, 2)
252
+ value_states = value_states.transpose(1, 2)
253
+
254
+ seq_dim = 1 if flash_attn_available else 2
255
+ kv_seq_len = key_states.shape[seq_dim]
256
+ if past_key_value is not None:
257
+ kv_seq_len += past_key_value[0].shape[seq_dim]
258
+ cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
259
+ query_states, key_states = apply_rotary_pos_emb(
260
+ query_states, key_states, cos, sin, position_ids, flash_attn_available
261
+ )
262
+
263
+ if past_key_value is not None:
264
+ # reuse k, v, self_attention
265
+ key_states = torch.cat([past_key_value[0], key_states], dim=seq_dim)
266
+ value_states = torch.cat([past_key_value[1], value_states], dim=seq_dim)
267
+
268
+ past_key_value = (key_states, value_states) if use_cache else None
269
+
270
+ if flash_attn_available:
271
+ attn_output = flash_attn_func(
272
+ query_states, key_states, value_states, dropout_p=0.0, causal=True
273
+ )
274
+ else:
275
+ attn_weights = torch.matmul(
276
+ query_states, key_states.transpose(2, 3)
277
+ ) / math.sqrt(self.head_dim)
278
+
279
+ if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
280
+ raise ValueError(
281
+ f"Attention weights should be of size {(bsz, self.num_heads, q_len, kv_seq_len)}, but is"
282
+ f" {attn_weights.size()}"
283
+ )
284
+
285
+ if attention_mask is not None:
286
+ if attention_mask.size() != (bsz, 1, q_len, kv_seq_len):
287
+ raise ValueError(
288
+ f"Attention mask should be of size {(bsz, 1, q_len, kv_seq_len)}, but is"
289
+ f"{attention_mask.size()}"
290
+ )
291
+ attn_weights = attn_weights + attention_mask
292
+ dtype_min = torch.tensor(
293
+ torch.finfo(attn_weights.dtype).min,
294
+ device=attn_weights.device,
295
+ dtype=attn_weights.dtype,
296
+ )
297
+ attn_weights = torch.max(attn_weights, dtype_min)
298
+
299
+ # upcast attention to fp32
300
+ attn_weights = nn.functional.softmax(
301
+ attn_weights, dim=-1, dtype=torch.float32
302
+ ).to(query_states.dtype)
303
+ attn_output = torch.matmul(attn_weights, value_states)
304
+
305
+ if attn_output.size() != (bsz, self.num_heads, q_len, self.head_dim):
306
+ raise ValueError(
307
+ f"`attn_output` should be of size {(bsz, self.num_heads, q_len, self.head_dim)}, but is"
308
+ f" {attn_output.size()}"
309
+ )
310
+
311
+ if not flash_attn_available:
312
+ attn_output = attn_output.transpose(1, 2)
313
+
314
+ attn_output = attn_output.reshape(bsz, q_len, self.hidden_size)
315
+
316
+ attn_output = self.o_proj(attn_output)
317
+
318
+ if not output_attentions:
319
+ attn_weights = None
320
+
321
+ return attn_output, attn_weights, past_key_value
322
+
323
+
324
+ class YiDecoderLayer(nn.Module):
325
+ def __init__(self, config: YiConfig):
326
+ super().__init__()
327
+
328
+ self.hidden_size = config.hidden_size
329
+ self.self_attn = YiAttention(config=config)
330
+ self.mlp = YiMLP(
331
+ hidden_size=self.hidden_size,
332
+ intermediate_size=config.intermediate_size,
333
+ hidden_act=config.hidden_act,
334
+ )
335
+
336
+ self.ln1 = YiRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
337
+ self.ln2 = YiRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
338
+
339
+ def forward(
340
+ self,
341
+ hidden_states: torch.Tensor,
342
+ attention_mask: Optional[torch.Tensor] = None,
343
+ position_ids: Optional[torch.LongTensor] = None,
344
+ past_key_value: Optional[Tuple[torch.Tensor]] = None,
345
+ output_attentions: Optional[bool] = False,
346
+ use_cache: Optional[bool] = False,
347
+ ) -> Tuple[
348
+ torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]
349
+ ]:
350
+ """
351
+ Args:
352
+ hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)`
353
+ attention_mask (`torch.FloatTensor`, *optional*): attention mask of size
354
+ `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values.
355
+ output_attentions (`bool`, *optional*):
356
+ Whether or not to return the attentions tensors of all attention layers. See `attentions` under
357
+ returned tensors for more detail.
358
+ use_cache (`bool`, *optional*):
359
+ If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding
360
+ (see `past_key_values`).
361
+ past_key_value (`Tuple(torch.FloatTensor)`, *optional*): cached past key and value projection states
362
+ """
363
+
364
+ residual = hidden_states
365
+
366
+ hidden_states = self.ln1(hidden_states)
367
+
368
+ # Self Attention
369
+ hidden_states, self_attn_weights, present_key_value = self.self_attn(
370
+ hidden_states=hidden_states,
371
+ attention_mask=attention_mask,
372
+ position_ids=position_ids,
373
+ past_key_value=past_key_value,
374
+ output_attentions=output_attentions,
375
+ use_cache=use_cache,
376
+ )
377
+ hidden_states = residual + hidden_states
378
+
379
+ # Fully Connected
380
+ residual = hidden_states
381
+ hidden_states = self.ln2(hidden_states)
382
+ hidden_states = self.mlp(hidden_states)
383
+ hidden_states = residual + hidden_states
384
+
385
+ outputs = (hidden_states,)
386
+
387
+ if output_attentions:
388
+ outputs += (self_attn_weights,)
389
+
390
+ if use_cache:
391
+ outputs += (present_key_value,)
392
+
393
+ return outputs
394
+
395
+
396
+ Yi_START_DOCSTRING = r"""
397
+ This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
398
+ library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
399
+ etc.)
400
+
401
+ This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
402
+ Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
403
+ and behavior.
404
+
405
+ Parameters:
406
+ config ([`YiConfig`]):
407
+ Model configuration class with all the parameters of the model. Initializing with a config file does not
408
+ load the weights associated with the model, only the configuration. Check out the
409
+ [`~PreTrainedModel.from_pretrained`] method to load the model weights.
410
+ """
411
+
412
+
413
+ @add_start_docstrings(
414
+ "The bare Yi Model outputting raw hidden-states without any specific head on top.",
415
+ Yi_START_DOCSTRING,
416
+ )
417
+ class YiPreTrainedModel(PreTrainedModel):
418
+ config_class = YiConfig
419
+ base_model_prefix = "model"
420
+ supports_gradient_checkpointing = True
421
+ _no_split_modules = ["YiDecoderLayer"]
422
+ _skip_keys_device_placement = "past_key_values"
423
+
424
+ def _init_weights(self, module):
425
+ std = self.config.initializer_range
426
+ if isinstance(module, nn.Linear):
427
+ module.weight.data.normal_(mean=0.0, std=std)
428
+ if module.bias is not None:
429
+ module.bias.data.zero_()
430
+ elif isinstance(module, nn.Embedding):
431
+ module.weight.data.normal_(mean=0.0, std=std)
432
+ if module.padding_idx is not None:
433
+ module.weight.data[module.padding_idx].zero_()
434
+
435
+ def _set_gradient_checkpointing(self, module, value=False):
436
+ if isinstance(module, YiModel):
437
+ module.gradient_checkpointing = value
438
+
439
+
440
+ Yi_INPUTS_DOCSTRING = r"""
441
+ Args:
442
+ input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
443
+ Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
444
+ it.
445
+
446
+ Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
447
+ [`PreTrainedTokenizer.__call__`] for details.
448
+
449
+ [What are input IDs?](../glossary#input-ids)
450
+ attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
451
+ Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
452
+
453
+ - 1 for tokens that are **not masked**,
454
+ - 0 for tokens that are **masked**.
455
+
456
+ [What are attention masks?](../glossary#attention-mask)
457
+
458
+ Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
459
+ [`PreTrainedTokenizer.__call__`] for details.
460
+
461
+ If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
462
+ `past_key_values`).
463
+
464
+ If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
465
+ and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
466
+ information on the default strategy.
467
+
468
+ - 1 indicates the head is **not masked**,
469
+ - 0 indicates the head is **masked**.
470
+ position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
471
+ Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
472
+ config.n_positions - 1]`.
473
+
474
+ [What are position IDs?](../glossary#position-ids)
475
+ past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
476
+ Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
477
+ `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
478
+ `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
479
+
480
+ Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
481
+ blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.
482
+
483
+ If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
484
+ don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
485
+ `decoder_input_ids` of shape `(batch_size, sequence_length)`.
486
+ inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
487
+ Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
488
+ is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
489
+ model's internal embedding lookup matrix.
490
+ use_cache (`bool`, *optional*):
491
+ If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
492
+ `past_key_values`).
493
+ output_attentions (`bool`, *optional*):
494
+ Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
495
+ tensors for more detail.
496
+ output_hidden_states (`bool`, *optional*):
497
+ Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
498
+ more detail.
499
+ return_dict (`bool`, *optional*):
500
+ Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
501
+ """
502
+
503
+
504
+ @add_start_docstrings(
505
+ "The bare Yi Model outputting raw hidden-states without any specific head on top.",
506
+ Yi_START_DOCSTRING,
507
+ )
508
+ class YiModel(YiPreTrainedModel):
509
+ """
510
+ Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`YiDecoderLayer`]
511
+
512
+ Args:
513
+ config: YiConfig
514
+ """
515
+
516
+ def __init__(self, config: YiConfig):
517
+ super().__init__(config)
518
+ self.padding_idx = config.pad_token_id
519
+ self.vocab_size = config.vocab_size
520
+
521
+ self.embed_tokens = nn.Embedding(
522
+ config.vocab_size, config.hidden_size, self.padding_idx
523
+ )
524
+ self.layers = nn.ModuleList(
525
+ [YiDecoderLayer(config) for _ in range(config.num_hidden_layers)]
526
+ )
527
+
528
+ self.norm = YiRMSNorm(config.hidden_size, eps=config.rms_norm_eps)
529
+
530
+ self.gradient_checkpointing = False
531
+ # Initialize weights and apply final processing
532
+ self.post_init()
533
+
534
+ def get_input_embeddings(self):
535
+ return self.embed_tokens
536
+
537
+ def set_input_embeddings(self, value):
538
+ self.embed_tokens = value
539
+
540
+ # Copied from transformers.models.bart.modeling_bart.BartDecoder._prepare_decoder_attention_mask
541
+ def _prepare_decoder_attention_mask(
542
+ self, attention_mask, input_ids, inputs_embeds, past_key_values_length
543
+ ):
544
+ input_shape = input_ids.shape
545
+ # create causal mask
546
+ # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
547
+ combined_attention_mask = None
548
+ if input_shape[-1] > 1:
549
+ combined_attention_mask = _make_causal_mask(
550
+ input_shape,
551
+ inputs_embeds.dtype,
552
+ device=inputs_embeds.device,
553
+ past_key_values_length=past_key_values_length,
554
+ )
555
+
556
+ if attention_mask is not None:
557
+ # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len]
558
+ expanded_attn_mask = _expand_mask(
559
+ attention_mask, inputs_embeds.dtype, tgt_len=input_shape[-1]
560
+ ).to(inputs_embeds.device)
561
+ combined_attention_mask = (
562
+ expanded_attn_mask
563
+ if combined_attention_mask is None
564
+ else expanded_attn_mask + combined_attention_mask
565
+ )
566
+
567
+ return combined_attention_mask
568
+
569
+ @add_start_docstrings_to_model_forward(Yi_INPUTS_DOCSTRING)
570
+ def forward(
571
+ self,
572
+ input_ids: torch.LongTensor = None,
573
+ attention_mask: Optional[torch.Tensor] = None,
574
+ position_ids: Optional[torch.LongTensor] = None,
575
+ past_key_values: Optional[List[torch.FloatTensor]] = None,
576
+ inputs_embeds: Optional[torch.FloatTensor] = None,
577
+ use_cache: Optional[bool] = None,
578
+ output_attentions: Optional[bool] = None,
579
+ output_hidden_states: Optional[bool] = None,
580
+ return_dict: Optional[bool] = None,
581
+ ) -> Union[Tuple, BaseModelOutputWithPast]:
582
+ output_attentions = (
583
+ output_attentions
584
+ if output_attentions is not None
585
+ else self.config.output_attentions
586
+ )
587
+ output_hidden_states = (
588
+ output_hidden_states
589
+ if output_hidden_states is not None
590
+ else self.config.output_hidden_states
591
+ )
592
+ use_cache = use_cache if use_cache is not None else self.config.use_cache
593
+
594
+ return_dict = (
595
+ return_dict if return_dict is not None else self.config.use_return_dict
596
+ )
597
+
598
+ # retrieve input_ids and inputs_embeds
599
+ if input_ids is not None and inputs_embeds is not None:
600
+ raise ValueError(
601
+ "You cannot specify both decoder_input_ids and decoder_inputs_embeds at the same time"
602
+ )
603
+ elif input_ids is not None:
604
+ batch_size, seq_length = input_ids.shape
605
+ elif inputs_embeds is not None:
606
+ batch_size, seq_length, _ = inputs_embeds.shape
607
+ else:
608
+ raise ValueError(
609
+ "You have to specify either decoder_input_ids or decoder_inputs_embeds"
610
+ )
611
+
612
+ seq_length_with_past = seq_length
613
+ past_key_values_length = 0
614
+
615
+ if past_key_values is not None:
616
+ past_key_values_length = past_key_values[0][0].shape[2]
617
+ seq_length_with_past = seq_length_with_past + past_key_values_length
618
+
619
+ if position_ids is None:
620
+ device = input_ids.device if input_ids is not None else inputs_embeds.device
621
+ position_ids = torch.arange(
622
+ past_key_values_length,
623
+ seq_length + past_key_values_length,
624
+ dtype=torch.long,
625
+ device=device,
626
+ )
627
+ position_ids = position_ids.unsqueeze(0).view(-1, seq_length)
628
+ else:
629
+ position_ids = position_ids.view(-1, seq_length).long()
630
+
631
+ if inputs_embeds is None:
632
+ inputs_embeds = self.embed_tokens(input_ids)
633
+
634
+ if not is_flash_attn_available():
635
+ # embed positions
636
+ if attention_mask is None:
637
+ attention_mask = torch.ones(
638
+ (batch_size, seq_length_with_past),
639
+ dtype=torch.bool,
640
+ device=inputs_embeds.device,
641
+ )
642
+ attention_mask = self._prepare_decoder_attention_mask(
643
+ attention_mask,
644
+ input_ids,
645
+ inputs_embeds,
646
+ past_key_values_length,
647
+ )
648
+ else:
649
+ attention_mask = None
650
+
651
+ hidden_states = inputs_embeds
652
+ if self.gradient_checkpointing and self.training:
653
+ if use_cache:
654
+ logger.warning_once(
655
+ "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..."
656
+ )
657
+ use_cache = False
658
+
659
+ # decoder layers
660
+ all_hidden_states = () if output_hidden_states else None
661
+ all_self_attns = () if output_attentions else None
662
+ next_decoder_cache = () if use_cache else None
663
+
664
+ for idx, decoder_layer in enumerate(self.layers):
665
+ if output_hidden_states:
666
+ all_hidden_states += (hidden_states,)
667
+
668
+ past_key_value = (
669
+ past_key_values[idx] if past_key_values is not None else None
670
+ )
671
+
672
+ if self.gradient_checkpointing and self.training:
673
+
674
+ def create_custom_forward(module):
675
+ def custom_forward(*inputs):
676
+ # None for past_key_value
677
+ return module(*inputs, past_key_value, output_attentions)
678
+
679
+ return custom_forward
680
+
681
+ layer_outputs = torch.utils.checkpoint.checkpoint(
682
+ create_custom_forward(decoder_layer),
683
+ hidden_states,
684
+ attention_mask,
685
+ position_ids,
686
+ )
687
+ else:
688
+ layer_outputs = decoder_layer(
689
+ hidden_states,
690
+ attention_mask=attention_mask,
691
+ position_ids=position_ids,
692
+ past_key_value=past_key_value,
693
+ output_attentions=output_attentions,
694
+ use_cache=use_cache,
695
+ )
696
+
697
+ hidden_states = layer_outputs[0]
698
+
699
+ if use_cache:
700
+ next_decoder_cache += (layer_outputs[2 if output_attentions else 1],)
701
+
702
+ if output_attentions:
703
+ all_self_attns += (layer_outputs[1],)
704
+
705
+ hidden_states = self.norm(hidden_states)
706
+ # add hidden states from the last decoder layer
707
+ if output_hidden_states:
708
+ all_hidden_states += (hidden_states,)
709
+
710
+ next_cache = next_decoder_cache if use_cache else None
711
+ if not return_dict:
712
+ return tuple(
713
+ v
714
+ for v in [hidden_states, next_cache, all_hidden_states, all_self_attns]
715
+ if v is not None
716
+ )
717
+ return BaseModelOutputWithPast(
718
+ last_hidden_state=hidden_states,
719
+ past_key_values=next_cache,
720
+ hidden_states=all_hidden_states,
721
+ attentions=all_self_attns,
722
+ )
723
+
724
+
725
+ class YiForCausalLM(YiPreTrainedModel):
726
+ _tied_weights_keys = ["lm_head.weight"]
727
+
728
+ def __init__(self, config):
729
+ super().__init__(config)
730
+ self.model = YiModel(config)
731
+
732
+ self.lm_head = nn.Linear(config.hidden_size, config.vocab_size, bias=False)
733
+
734
+ # Initialize weights and apply final processing
735
+ self.post_init()
736
+
737
+ def get_input_embeddings(self):
738
+ return self.model.embed_tokens
739
+
740
+ def set_input_embeddings(self, value):
741
+ self.model.embed_tokens = value
742
+
743
+ def get_output_embeddings(self):
744
+ return self.lm_head
745
+
746
+ def set_output_embeddings(self, new_embeddings):
747
+ self.lm_head = new_embeddings
748
+
749
+ def set_decoder(self, decoder):
750
+ self.model = decoder
751
+
752
+ def get_decoder(self):
753
+ return self.model
754
+
755
+ @add_start_docstrings_to_model_forward(Yi_INPUTS_DOCSTRING)
756
+ @replace_return_docstrings(
757
+ output_type=CausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC
758
+ )
759
+ def forward(
760
+ self,
761
+ input_ids: torch.LongTensor = None,
762
+ attention_mask: Optional[torch.Tensor] = None,
763
+ position_ids: Optional[torch.LongTensor] = None,
764
+ past_key_values: Optional[List[torch.FloatTensor]] = None,
765
+ inputs_embeds: Optional[torch.FloatTensor] = None,
766
+ labels: Optional[torch.LongTensor] = None,
767
+ use_cache: Optional[bool] = None,
768
+ output_attentions: Optional[bool] = None,
769
+ output_hidden_states: Optional[bool] = None,
770
+ return_dict: Optional[bool] = None,
771
+ ) -> Union[Tuple, CausalLMOutputWithPast]:
772
+ r"""
773
+ Args:
774
+ labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
775
+ Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
776
+ config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
777
+ (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
778
+
779
+ Returns:
780
+
781
+ Example:
782
+
783
+ ```python
784
+ >>> from transformers import AutoTokenizer, YiForCausalLM
785
+
786
+ >>> model = YiForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
787
+ >>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
788
+
789
+ >>> prompt = "Hey, are you conscious? Can you talk to me?"
790
+ >>> inputs = tokenizer(prompt, return_tensors="pt")
791
+
792
+ >>> # Generate
793
+ >>> generate_ids = model.generate(inputs.input_ids, max_length=30)
794
+ >>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
795
+ "Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
796
+ ```"""
797
+
798
+ output_attentions = (
799
+ output_attentions
800
+ if output_attentions is not None
801
+ else self.config.output_attentions
802
+ )
803
+ output_hidden_states = (
804
+ output_hidden_states
805
+ if output_hidden_states is not None
806
+ else self.config.output_hidden_states
807
+ )
808
+ return_dict = (
809
+ return_dict if return_dict is not None else self.config.use_return_dict
810
+ )
811
+
812
+ # decoder outputs consists of (dec_features, layer_state, dec_hidden, dec_attn)
813
+ outputs = self.model(
814
+ input_ids=input_ids,
815
+ attention_mask=attention_mask,
816
+ position_ids=position_ids,
817
+ past_key_values=past_key_values,
818
+ inputs_embeds=inputs_embeds,
819
+ use_cache=use_cache,
820
+ output_attentions=output_attentions,
821
+ output_hidden_states=output_hidden_states,
822
+ return_dict=return_dict,
823
+ )
824
+
825
+ hidden_states = outputs[0]
826
+ logits = self.lm_head(hidden_states)
827
+
828
+ loss = None
829
+ if labels is not None:
830
+ # Shift so that tokens < n predict n
831
+ shift_logits = logits[..., :-1, :].contiguous()
832
+ shift_labels = labels[..., 1:].contiguous()
833
+ # Flatten the tokens
834
+ loss_fct = CrossEntropyLoss()
835
+ shift_logits = shift_logits.view(-1, self.config.vocab_size)
836
+ shift_labels = shift_labels.view(-1)
837
+ # Enable model parallelism
838
+ shift_labels = shift_labels.to(shift_logits.device)
839
+ loss = loss_fct(shift_logits, shift_labels)
840
+
841
+ if not return_dict:
842
+ output = (logits,) + outputs[1:]
843
+ return (loss,) + output if loss is not None else output
844
+
845
+ return CausalLMOutputWithPast(
846
+ loss=loss,
847
+ logits=logits,
848
+ past_key_values=outputs.past_key_values,
849
+ hidden_states=outputs.hidden_states,
850
+ attentions=outputs.attentions,
851
+ )
852
+
853
+ def prepare_inputs_for_generation(
854
+ self,
855
+ input_ids,
856
+ past_key_values=None,
857
+ attention_mask=None,
858
+ inputs_embeds=None,
859
+ **kwargs,
860
+ ):
861
+ if past_key_values:
862
+ input_ids = input_ids[:, -1:]
863
+
864
+ position_ids = kwargs.get("position_ids", None)
865
+ if attention_mask is not None and position_ids is None:
866
+ # create position_ids on the fly for batch generation
867
+ position_ids = attention_mask.long().cumsum(-1) - 1
868
+ position_ids.masked_fill_(attention_mask == 0, 1)
869
+ if past_key_values:
870
+ position_ids = position_ids[:, -1].unsqueeze(-1)
871
+
872
+ # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
873
+ if inputs_embeds is not None and past_key_values is None:
874
+ model_inputs = {"inputs_embeds": inputs_embeds}
875
+ else:
876
+ model_inputs = {"input_ids": input_ids}
877
+
878
+ model_inputs.update(
879
+ {
880
+ "position_ids": position_ids,
881
+ "past_key_values": past_key_values,
882
+ "use_cache": kwargs.get("use_cache"),
883
+ "attention_mask": attention_mask,
884
+ }
885
+ )
886
+ return model_inputs
887
+
888
+ @staticmethod
889
+ def _reorder_cache(past_key_values, beam_idx):
890
+ reordered_past = ()
891
+ for layer_past in past_key_values:
892
+ reordered_past += (
893
+ tuple(
894
+ past_state.index_select(0, beam_idx.to(past_state.device))
895
+ for past_state in layer_past
896
+ ),
897
+ )
898
+ return reordered_past
899
+
900
+
901
+ @add_start_docstrings(
902
+ """
903
+ The Yi Model transformer with a sequence classification head on top (linear layer).
904
+
905
+ [`YiForSequenceClassification`] uses the last token in order to do the classification, as other causal models
906
+ (e.g. GPT-2) do.
907
+
908
+ Since it does classification on the last token, it requires to know the position of the last token. If a
909
+ `pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If
910
+ no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the
911
+ padding tokens when `inputs_embeds` are passed instead of `input_ids`, it does the same (take the last value in
912
+ each row of the batch).
913
+ """,
914
+ Yi_START_DOCSTRING,
915
+ )
916
+ class YiForSequenceClassification(YiPreTrainedModel):
917
+ def __init__(self, config):
918
+ super().__init__(config)
919
+ self.num_labels = config.num_labels
920
+ self.model = YiModel(config)
921
+ self.score = nn.Linear(config.hidden_size, self.num_labels, bias=False)
922
+
923
+ # Initialize weights and apply final processing
924
+ self.post_init()
925
+
926
+ def get_input_embeddings(self):
927
+ return self.model.embed_tokens
928
+
929
+ def set_input_embeddings(self, value):
930
+ self.model.embed_tokens = value
931
+
932
+ @add_start_docstrings_to_model_forward(Yi_INPUTS_DOCSTRING)
933
+ def forward(
934
+ self,
935
+ input_ids: torch.LongTensor = None,
936
+ attention_mask: Optional[torch.Tensor] = None,
937
+ position_ids: Optional[torch.LongTensor] = None,
938
+ past_key_values: Optional[List[torch.FloatTensor]] = None,
939
+ inputs_embeds: Optional[torch.FloatTensor] = None,
940
+ labels: Optional[torch.LongTensor] = None,
941
+ use_cache: Optional[bool] = None,
942
+ output_attentions: Optional[bool] = None,
943
+ output_hidden_states: Optional[bool] = None,
944
+ return_dict: Optional[bool] = None,
945
+ ) -> Union[Tuple, SequenceClassifierOutputWithPast]:
946
+ r"""
947
+ labels (`torch.LongTensor` of shape `(batch_size,)`, *optional*):
948
+ Labels for computing the sequence classification/regression loss. Indices should be in `[0, ...,
949
+ config.num_labels - 1]`. If `config.num_labels == 1` a regression loss is computed (Mean-Square loss), If
950
+ `config.num_labels > 1` a classification loss is computed (Cross-Entropy).
951
+ """
952
+ return_dict = (
953
+ return_dict if return_dict is not None else self.config.use_return_dict
954
+ )
955
+
956
+ transformer_outputs = self.model(
957
+ input_ids,
958
+ attention_mask=attention_mask,
959
+ position_ids=position_ids,
960
+ past_key_values=past_key_values,
961
+ inputs_embeds=inputs_embeds,
962
+ use_cache=use_cache,
963
+ output_attentions=output_attentions,
964
+ output_hidden_states=output_hidden_states,
965
+ return_dict=return_dict,
966
+ )
967
+ hidden_states = transformer_outputs[0]
968
+ logits = self.score(hidden_states)
969
+
970
+ if input_ids is not None:
971
+ batch_size = input_ids.shape[0]
972
+ else:
973
+ batch_size = inputs_embeds.shape[0]
974
+
975
+ if self.config.pad_token_id is None and batch_size != 1:
976
+ raise ValueError(
977
+ "Cannot handle batch sizes > 1 if no padding token is defined."
978
+ )
979
+ if self.config.pad_token_id is None:
980
+ sequence_lengths = -1
981
+ else:
982
+ if input_ids is not None:
983
+ sequence_lengths = (
984
+ torch.eq(input_ids, self.config.pad_token_id).long().argmax(-1) - 1
985
+ ).to(logits.device)
986
+ else:
987
+ sequence_lengths = -1
988
+
989
+ pooled_logits = logits[
990
+ torch.arange(batch_size, device=logits.device), sequence_lengths
991
+ ]
992
+
993
+ loss = None
994
+ if labels is not None:
995
+ labels = labels.to(logits.device)
996
+ if self.config.problem_type is None:
997
+ if self.num_labels == 1:
998
+ self.config.problem_type = "regression"
999
+ elif self.num_labels > 1 and (
1000
+ labels.dtype == torch.long or labels.dtype == torch.int
1001
+ ):
1002
+ self.config.problem_type = "single_label_classification"
1003
+ else:
1004
+ self.config.problem_type = "multi_label_classification"
1005
+
1006
+ if self.config.problem_type == "regression":
1007
+ loss_fct = MSELoss()
1008
+ if self.num_labels == 1:
1009
+ loss = loss_fct(pooled_logits.squeeze(), labels.squeeze())
1010
+ else:
1011
+ loss = loss_fct(pooled_logits, labels)
1012
+ elif self.config.problem_type == "single_label_classification":
1013
+ loss_fct = CrossEntropyLoss()
1014
+ loss = loss_fct(
1015
+ pooled_logits.view(-1, self.num_labels), labels.view(-1)
1016
+ )
1017
+ elif self.config.problem_type == "multi_label_classification":
1018
+ loss_fct = BCEWithLogitsLoss()
1019
+ loss = loss_fct(pooled_logits, labels)
1020
+ if not return_dict:
1021
+ output = (pooled_logits,) + transformer_outputs[1:]
1022
+ return ((loss,) + output) if loss is not None else output
1023
+
1024
+ return SequenceClassifierOutputWithPast(
1025
+ loss=loss,
1026
+ logits=pooled_logits,
1027
+ past_key_values=transformer_outputs.past_key_values,
1028
+ hidden_states=transformer_outputs.hidden_states,
1029
+ attentions=transformer_outputs.attentions,
1030
+ )
pytorch_model-00001-of-00002.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7335422e13c29396b729da0337edd6280f0ccee4b714a59051547b0f5d802d13
3
+ size 9943124771
pytorch_model-00002-of-00002.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:145fb5246919be8da75f5ed01b02fc45ce8dea0cba43d1e095892c98be81e1ff
3
+ size 2179045205
pytorch_model.bin.index.json ADDED
@@ -0,0 +1,298 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 12122071040
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "pytorch_model-00002-of-00002.bin",
7
+ "model.embed_tokens.weight": "pytorch_model-00001-of-00002.bin",
8
+ "model.layers.0.ln1.weight": "pytorch_model-00001-of-00002.bin",
9
+ "model.layers.0.ln2.weight": "pytorch_model-00001-of-00002.bin",
10
+ "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
11
+ "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
12
+ "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
13
+ "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
14
+ "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
15
+ "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
16
+ "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
17
+ "model.layers.1.ln1.weight": "pytorch_model-00001-of-00002.bin",
18
+ "model.layers.1.ln2.weight": "pytorch_model-00001-of-00002.bin",
19
+ "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
20
+ "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
21
+ "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
22
+ "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
23
+ "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
24
+ "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
25
+ "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
26
+ "model.layers.10.ln1.weight": "pytorch_model-00001-of-00002.bin",
27
+ "model.layers.10.ln2.weight": "pytorch_model-00001-of-00002.bin",
28
+ "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
29
+ "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
30
+ "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
31
+ "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
32
+ "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
33
+ "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
34
+ "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
35
+ "model.layers.11.ln1.weight": "pytorch_model-00001-of-00002.bin",
36
+ "model.layers.11.ln2.weight": "pytorch_model-00001-of-00002.bin",
37
+ "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
38
+ "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
39
+ "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
40
+ "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
41
+ "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
42
+ "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
43
+ "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
44
+ "model.layers.12.ln1.weight": "pytorch_model-00001-of-00002.bin",
45
+ "model.layers.12.ln2.weight": "pytorch_model-00001-of-00002.bin",
46
+ "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
47
+ "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
48
+ "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
49
+ "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
50
+ "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
51
+ "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
52
+ "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
53
+ "model.layers.13.ln1.weight": "pytorch_model-00001-of-00002.bin",
54
+ "model.layers.13.ln2.weight": "pytorch_model-00001-of-00002.bin",
55
+ "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
56
+ "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
57
+ "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
58
+ "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
59
+ "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
60
+ "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
61
+ "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
62
+ "model.layers.14.ln1.weight": "pytorch_model-00001-of-00002.bin",
63
+ "model.layers.14.ln2.weight": "pytorch_model-00001-of-00002.bin",
64
+ "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
65
+ "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
66
+ "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
67
+ "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
68
+ "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
69
+ "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
70
+ "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
71
+ "model.layers.15.ln1.weight": "pytorch_model-00001-of-00002.bin",
72
+ "model.layers.15.ln2.weight": "pytorch_model-00001-of-00002.bin",
73
+ "model.layers.15.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
74
+ "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
75
+ "model.layers.15.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
76
+ "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
77
+ "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
78
+ "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
79
+ "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
80
+ "model.layers.16.ln1.weight": "pytorch_model-00001-of-00002.bin",
81
+ "model.layers.16.ln2.weight": "pytorch_model-00001-of-00002.bin",
82
+ "model.layers.16.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
83
+ "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
84
+ "model.layers.16.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
85
+ "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
86
+ "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
87
+ "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
88
+ "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
89
+ "model.layers.17.ln1.weight": "pytorch_model-00001-of-00002.bin",
90
+ "model.layers.17.ln2.weight": "pytorch_model-00001-of-00002.bin",
91
+ "model.layers.17.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
92
+ "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
93
+ "model.layers.17.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
94
+ "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
95
+ "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
96
+ "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
97
+ "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
98
+ "model.layers.18.ln1.weight": "pytorch_model-00001-of-00002.bin",
99
+ "model.layers.18.ln2.weight": "pytorch_model-00001-of-00002.bin",
100
+ "model.layers.18.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
101
+ "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
102
+ "model.layers.18.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
103
+ "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
104
+ "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
105
+ "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
106
+ "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
107
+ "model.layers.19.ln1.weight": "pytorch_model-00001-of-00002.bin",
108
+ "model.layers.19.ln2.weight": "pytorch_model-00001-of-00002.bin",
109
+ "model.layers.19.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
110
+ "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
111
+ "model.layers.19.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
112
+ "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
113
+ "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
114
+ "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
115
+ "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
116
+ "model.layers.2.ln1.weight": "pytorch_model-00001-of-00002.bin",
117
+ "model.layers.2.ln2.weight": "pytorch_model-00001-of-00002.bin",
118
+ "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
119
+ "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
120
+ "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
121
+ "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
122
+ "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
123
+ "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
124
+ "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
125
+ "model.layers.20.ln1.weight": "pytorch_model-00001-of-00002.bin",
126
+ "model.layers.20.ln2.weight": "pytorch_model-00001-of-00002.bin",
127
+ "model.layers.20.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
128
+ "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
129
+ "model.layers.20.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
130
+ "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
131
+ "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
132
+ "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
133
+ "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
134
+ "model.layers.21.ln1.weight": "pytorch_model-00001-of-00002.bin",
135
+ "model.layers.21.ln2.weight": "pytorch_model-00001-of-00002.bin",
136
+ "model.layers.21.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
137
+ "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
138
+ "model.layers.21.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
139
+ "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
140
+ "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
141
+ "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
142
+ "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
143
+ "model.layers.22.ln1.weight": "pytorch_model-00001-of-00002.bin",
144
+ "model.layers.22.ln2.weight": "pytorch_model-00001-of-00002.bin",
145
+ "model.layers.22.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
146
+ "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
147
+ "model.layers.22.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
148
+ "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
149
+ "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
150
+ "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
151
+ "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
152
+ "model.layers.23.ln1.weight": "pytorch_model-00001-of-00002.bin",
153
+ "model.layers.23.ln2.weight": "pytorch_model-00001-of-00002.bin",
154
+ "model.layers.23.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
155
+ "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
156
+ "model.layers.23.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
157
+ "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
158
+ "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
159
+ "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
160
+ "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
161
+ "model.layers.24.ln1.weight": "pytorch_model-00001-of-00002.bin",
162
+ "model.layers.24.ln2.weight": "pytorch_model-00001-of-00002.bin",
163
+ "model.layers.24.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
164
+ "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
165
+ "model.layers.24.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
166
+ "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
167
+ "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
168
+ "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
169
+ "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
170
+ "model.layers.25.ln1.weight": "pytorch_model-00001-of-00002.bin",
171
+ "model.layers.25.ln2.weight": "pytorch_model-00001-of-00002.bin",
172
+ "model.layers.25.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
173
+ "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
174
+ "model.layers.25.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
175
+ "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
176
+ "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
177
+ "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
178
+ "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
179
+ "model.layers.26.ln1.weight": "pytorch_model-00001-of-00002.bin",
180
+ "model.layers.26.ln2.weight": "pytorch_model-00001-of-00002.bin",
181
+ "model.layers.26.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
182
+ "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
183
+ "model.layers.26.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
184
+ "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
185
+ "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
186
+ "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
187
+ "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
188
+ "model.layers.27.ln1.weight": "pytorch_model-00002-of-00002.bin",
189
+ "model.layers.27.ln2.weight": "pytorch_model-00002-of-00002.bin",
190
+ "model.layers.27.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
191
+ "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
192
+ "model.layers.27.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
193
+ "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
194
+ "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
195
+ "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
196
+ "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
197
+ "model.layers.28.ln1.weight": "pytorch_model-00002-of-00002.bin",
198
+ "model.layers.28.ln2.weight": "pytorch_model-00002-of-00002.bin",
199
+ "model.layers.28.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
200
+ "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
201
+ "model.layers.28.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
202
+ "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
203
+ "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
204
+ "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
205
+ "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
206
+ "model.layers.29.ln1.weight": "pytorch_model-00002-of-00002.bin",
207
+ "model.layers.29.ln2.weight": "pytorch_model-00002-of-00002.bin",
208
+ "model.layers.29.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
209
+ "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
210
+ "model.layers.29.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
211
+ "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
212
+ "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
213
+ "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
214
+ "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
215
+ "model.layers.3.ln1.weight": "pytorch_model-00001-of-00002.bin",
216
+ "model.layers.3.ln2.weight": "pytorch_model-00001-of-00002.bin",
217
+ "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
218
+ "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
219
+ "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
220
+ "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
221
+ "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
222
+ "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
223
+ "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
224
+ "model.layers.30.ln1.weight": "pytorch_model-00002-of-00002.bin",
225
+ "model.layers.30.ln2.weight": "pytorch_model-00002-of-00002.bin",
226
+ "model.layers.30.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
227
+ "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
228
+ "model.layers.30.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
229
+ "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
230
+ "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
231
+ "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
232
+ "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
233
+ "model.layers.31.ln1.weight": "pytorch_model-00002-of-00002.bin",
234
+ "model.layers.31.ln2.weight": "pytorch_model-00002-of-00002.bin",
235
+ "model.layers.31.mlp.down_proj.weight": "pytorch_model-00002-of-00002.bin",
236
+ "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00002-of-00002.bin",
237
+ "model.layers.31.mlp.up_proj.weight": "pytorch_model-00002-of-00002.bin",
238
+ "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00002-of-00002.bin",
239
+ "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00002-of-00002.bin",
240
+ "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00002-of-00002.bin",
241
+ "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00002-of-00002.bin",
242
+ "model.layers.4.ln1.weight": "pytorch_model-00001-of-00002.bin",
243
+ "model.layers.4.ln2.weight": "pytorch_model-00001-of-00002.bin",
244
+ "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
245
+ "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
246
+ "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
247
+ "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
248
+ "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
249
+ "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
250
+ "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
251
+ "model.layers.5.ln1.weight": "pytorch_model-00001-of-00002.bin",
252
+ "model.layers.5.ln2.weight": "pytorch_model-00001-of-00002.bin",
253
+ "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
254
+ "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
255
+ "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
256
+ "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
257
+ "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
258
+ "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
259
+ "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
260
+ "model.layers.6.ln1.weight": "pytorch_model-00001-of-00002.bin",
261
+ "model.layers.6.ln2.weight": "pytorch_model-00001-of-00002.bin",
262
+ "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
263
+ "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
264
+ "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
265
+ "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
266
+ "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
267
+ "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
268
+ "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
269
+ "model.layers.7.ln1.weight": "pytorch_model-00001-of-00002.bin",
270
+ "model.layers.7.ln2.weight": "pytorch_model-00001-of-00002.bin",
271
+ "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
272
+ "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
273
+ "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
274
+ "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
275
+ "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
276
+ "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
277
+ "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
278
+ "model.layers.8.ln1.weight": "pytorch_model-00001-of-00002.bin",
279
+ "model.layers.8.ln2.weight": "pytorch_model-00001-of-00002.bin",
280
+ "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
281
+ "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
282
+ "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
283
+ "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
284
+ "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
285
+ "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
286
+ "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
287
+ "model.layers.9.ln1.weight": "pytorch_model-00001-of-00002.bin",
288
+ "model.layers.9.ln2.weight": "pytorch_model-00001-of-00002.bin",
289
+ "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00002.bin",
290
+ "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00002.bin",
291
+ "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00002.bin",
292
+ "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00002.bin",
293
+ "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00002.bin",
294
+ "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00002.bin",
295
+ "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00002.bin",
296
+ "model.norm.weight": "pytorch_model-00002-of-00002.bin"
297
+ }
298
+ }
tokenization_yi.py ADDED
@@ -0,0 +1,255 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ from shutil import copyfile
3
+ from typing import Any, Dict, List, Optional, Tuple
4
+
5
+ import sentencepiece as spm
6
+ from transformers.tokenization_utils import AddedToken, PreTrainedTokenizer
7
+ from transformers.utils import logging
8
+
9
+ logger = logging.get_logger(__name__)
10
+
11
+ VOCAB_FILES_NAMES = {"vocab_file": "tokenizer.model"}
12
+
13
+ PRETRAINED_VOCAB_FILES_MAP = {
14
+ "vocab_file": {},
15
+ "tokenizer_file": {},
16
+ }
17
+ PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {}
18
+
19
+
20
+ class YiTokenizer(PreTrainedTokenizer):
21
+ """
22
+ Construct a Yi tokenizer. Based on byte-level Byte-Pair-Encoding.
23
+
24
+ Args:
25
+ vocab_file (`str`):
26
+ Path to the vocabulary file.
27
+ """
28
+
29
+ vocab_files_names = VOCAB_FILES_NAMES
30
+ pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
31
+ max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
32
+ model_input_names = ["input_ids", "attention_mask"]
33
+
34
+ def __init__(
35
+ self,
36
+ vocab_file,
37
+ unk_token="<unk>",
38
+ bos_token="<|startoftext|>",
39
+ eos_token="<|endoftext|>",
40
+ pad_token="<unk>",
41
+ sp_model_kwargs: Optional[Dict[str, Any]] = None,
42
+ add_bos_token=True,
43
+ add_eos_token=False,
44
+ clean_up_tokenization_spaces=False,
45
+ **kwargs,
46
+ ):
47
+ self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs
48
+ bos_token = (
49
+ AddedToken(bos_token, lstrip=False, rstrip=False)
50
+ if isinstance(bos_token, str)
51
+ else bos_token
52
+ )
53
+ eos_token = (
54
+ AddedToken(eos_token, lstrip=False, rstrip=False)
55
+ if isinstance(eos_token, str)
56
+ else eos_token
57
+ )
58
+ unk_token = (
59
+ AddedToken(unk_token, lstrip=False, rstrip=False)
60
+ if isinstance(unk_token, str)
61
+ else unk_token
62
+ )
63
+ pad_token = (
64
+ AddedToken(pad_token, lstrip=False, rstrip=False)
65
+ if isinstance(pad_token, str)
66
+ else pad_token
67
+ )
68
+ self.vocab_file = vocab_file
69
+ self.add_bos_token = add_bos_token
70
+ self.add_eos_token = add_eos_token
71
+ self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
72
+ self.sp_model.Load(vocab_file)
73
+ super().__init__(
74
+ bos_token=bos_token,
75
+ eos_token=eos_token,
76
+ unk_token=unk_token,
77
+ pad_token=pad_token,
78
+ add_bos_token=add_bos_token,
79
+ add_eos_token=add_eos_token,
80
+ sp_model_kwargs=self.sp_model_kwargs,
81
+ clean_up_tokenization_spaces=clean_up_tokenization_spaces,
82
+ **kwargs,
83
+ )
84
+
85
+ def __getstate__(self):
86
+ state = self.__dict__.copy()
87
+ state["sp_model"] = None
88
+ return state
89
+
90
+ def __setstate__(self, d):
91
+ self.__dict__ = d
92
+ self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
93
+ self.sp_model.Load(self.vocab_file)
94
+
95
+ @property
96
+ def vocab_size(self):
97
+ """Returns vocab size"""
98
+ return self.sp_model.get_piece_size()
99
+
100
+ def get_vocab(self):
101
+ """Returns vocab as a dict"""
102
+ vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
103
+ vocab.update(self.added_tokens_encoder)
104
+ return vocab
105
+
106
+ def _tokenize(self, text):
107
+ """Returns a tokenized string."""
108
+ return self.sp_model.encode(text, out_type=str)
109
+
110
+ def _convert_token_to_id(self, token):
111
+ """Converts a token (str) in an id using the vocab."""
112
+ return self.sp_model.piece_to_id(token)
113
+
114
+ def _convert_id_to_token(self, index):
115
+ """Converts an index (integer) in a token (str) using the vocab."""
116
+ token = self.sp_model.IdToPiece(index)
117
+ return token
118
+
119
+ def convert_tokens_to_string(self, tokens):
120
+ """Converts a sequence of tokens (string) in a single string."""
121
+ current_sub_tokens = []
122
+ out_string = ""
123
+ prev_is_special = False
124
+ for i, token in enumerate(tokens):
125
+ # make sure that special tokens are not decoded using sentencepiece model
126
+ if token in self.all_special_tokens:
127
+ if not prev_is_special and i != 0:
128
+ out_string += " "
129
+ out_string += self.sp_model.decode(current_sub_tokens) + token
130
+ prev_is_special = True
131
+ current_sub_tokens = []
132
+ else:
133
+ current_sub_tokens.append(token)
134
+ prev_is_special = False
135
+ out_string += self.sp_model.decode(current_sub_tokens)
136
+ return out_string
137
+
138
+ def save_vocabulary(
139
+ self, save_directory, filename_prefix: Optional[str] = None
140
+ ) -> Tuple[str]:
141
+ """
142
+ Save the vocabulary and special tokens file to a directory.
143
+
144
+ Args:
145
+ save_directory (`str`):
146
+ The directory in which to save the vocabulary.
147
+
148
+ Returns:
149
+ `Tuple(str)`: Paths to the files saved.
150
+ """
151
+ if not os.path.isdir(save_directory):
152
+ logger.error(f"Vocabulary path ({save_directory}) should be a directory")
153
+ return
154
+ out_vocab_file = os.path.join(
155
+ save_directory,
156
+ (filename_prefix + "-" if filename_prefix else "")
157
+ + VOCAB_FILES_NAMES["vocab_file"],
158
+ )
159
+
160
+ if os.path.abspath(self.vocab_file) != os.path.abspath(
161
+ out_vocab_file
162
+ ) and os.path.isfile(self.vocab_file):
163
+ copyfile(self.vocab_file, out_vocab_file)
164
+ elif not os.path.isfile(self.vocab_file):
165
+ with open(out_vocab_file, "wb") as fi:
166
+ content_spiece_model = self.sp_model.serialized_model_proto()
167
+ fi.write(content_spiece_model)
168
+
169
+ return (out_vocab_file,)
170
+
171
+ def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
172
+ bos_token_id = [self.bos_token_id] if self.add_bos_token else []
173
+ eos_token_id = [self.eos_token_id] if self.add_eos_token else []
174
+
175
+ output = bos_token_id + token_ids_0 + eos_token_id
176
+
177
+ if token_ids_1 is not None:
178
+ output = output + bos_token_id + token_ids_1 + eos_token_id
179
+
180
+ return output
181
+
182
+ def get_special_tokens_mask(
183
+ self,
184
+ token_ids_0: List[int],
185
+ token_ids_1: Optional[List[int]] = None,
186
+ already_has_special_tokens: bool = False,
187
+ ) -> List[int]:
188
+ """
189
+ Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
190
+ special tokens using the tokenizer `prepare_for_model` method.
191
+
192
+ Args:
193
+ token_ids_0 (`List[int]`):
194
+ List of IDs.
195
+ token_ids_1 (`List[int]`, *optional*):
196
+ Optional second list of IDs for sequence pairs.
197
+ already_has_special_tokens (`bool`, *optional*, defaults to `False`):
198
+ Whether or not the token list is already formatted with special tokens for the model.
199
+
200
+ Returns:
201
+ `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
202
+ """
203
+ if already_has_special_tokens:
204
+ return super().get_special_tokens_mask(
205
+ token_ids_0=token_ids_0,
206
+ token_ids_1=token_ids_1,
207
+ already_has_special_tokens=True,
208
+ )
209
+
210
+ bos_token_id = [1] if self.add_bos_token else []
211
+ eos_token_id = [1] if self.add_eos_token else []
212
+
213
+ if token_ids_1 is None:
214
+ return bos_token_id + ([0] * len(token_ids_0)) + eos_token_id
215
+ return (
216
+ bos_token_id
217
+ + ([0] * len(token_ids_0))
218
+ + eos_token_id
219
+ + bos_token_id
220
+ + ([0] * len(token_ids_1))
221
+ + eos_token_id
222
+ )
223
+
224
+ def create_token_type_ids_from_sequences(
225
+ self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
226
+ ) -> List[int]:
227
+ """
228
+ Creates a mask from the two sequences passed to be used in a sequence-pair classification task. An ALBERT
229
+ sequence pair mask has the following format:
230
+
231
+ ```
232
+ 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
233
+ | first sequence | second sequence |
234
+ ```
235
+
236
+ if token_ids_1 is None, only returns the first portion of the mask (0s).
237
+
238
+ Args:
239
+ token_ids_0 (`List[int]`):
240
+ List of ids.
241
+ token_ids_1 (`List[int]`, *optional*):
242
+ Optional second list of IDs for sequence pairs.
243
+
244
+ Returns:
245
+ `List[int]`: List of [token type IDs](../glossary#token-type-ids) according to the given sequence(s).
246
+ """
247
+ bos_token_id = [self.bos_token_id] if self.add_bos_token else []
248
+ eos_token_id = [self.eos_token_id] if self.add_eos_token else []
249
+
250
+ output = [0] * len(bos_token_id + token_ids_0 + eos_token_id)
251
+
252
+ if token_ids_1 is not None:
253
+ output += [1] * len(bos_token_id + token_ids_1 + eos_token_id)
254
+
255
+ return output
tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:386c49cf943d71aa110361135338c50e38beeff0a66593480421f37b319e1a39
3
+ size 1033105
tokenizer_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_map": {
3
+ "AutoTokenizer": ["tokenization_yi.YiTokenizer", null]
4
+ },
5
+ "add_bos_token": false,
6
+ "add_eos_token": false,
7
+ "model_max_length": 4096,
8
+ "tokenizer_class": "YiTokenizer"
9
+ }