lmdeploy-hint
#10
by
irexyc
- opened
- config.json +3 -6
- configuration_internlm.py +7 -8
- modeling_internlm.py +76 -167
config.json
CHANGED
@@ -1,4 +1,5 @@
|
|
1 |
{
|
|
|
2 |
"architectures": [
|
3 |
"InternLMForCausalLM"
|
4 |
],
|
@@ -19,7 +20,7 @@
|
|
19 |
"num_attention_heads": 40,
|
20 |
"num_hidden_layers": 60,
|
21 |
"num_key_value_heads": 40,
|
22 |
-
"pad_token_id":
|
23 |
"pretraining_tp": 1,
|
24 |
"rms_norm_eps": 1e-06,
|
25 |
"rope_scaling": null,
|
@@ -28,9 +29,5 @@
|
|
28 |
"torch_dtype": "float16",
|
29 |
"transformers_version": "4.33.1",
|
30 |
"use_cache": false,
|
31 |
-
"vocab_size": 103168
|
32 |
-
"rotary": {
|
33 |
-
"base": 10000,
|
34 |
-
"type": "dynamic"
|
35 |
-
}
|
36 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "/nvme/shared_data/InternLM/20B/internlm-20b-chat",
|
3 |
"architectures": [
|
4 |
"InternLMForCausalLM"
|
5 |
],
|
|
|
20 |
"num_attention_heads": 40,
|
21 |
"num_hidden_layers": 60,
|
22 |
"num_key_value_heads": 40,
|
23 |
+
"pad_token_id": 0,
|
24 |
"pretraining_tp": 1,
|
25 |
"rms_norm_eps": 1e-06,
|
26 |
"rope_scaling": null,
|
|
|
29 |
"torch_dtype": "float16",
|
30 |
"transformers_version": "4.33.1",
|
31 |
"use_cache": false,
|
32 |
+
"vocab_size": 103168
|
|
|
|
|
|
|
|
|
33 |
}
|
configuration_internlm.py
CHANGED
@@ -19,8 +19,9 @@
|
|
19 |
# limitations under the License.
|
20 |
""" InternLM model configuration"""
|
21 |
|
22 |
-
from transformers.configuration_utils import PretrainedConfig
|
23 |
from transformers.utils import logging
|
|
|
|
|
24 |
|
25 |
logger = logging.get_logger(__name__)
|
26 |
|
@@ -29,9 +30,9 @@ INTERNLM_PRETRAINED_CONFIG_ARCHIVE_MAP = {}
|
|
29 |
|
30 |
class InternLMConfig(PretrainedConfig):
|
31 |
r"""
|
32 |
-
This is the configuration class to store the configuration of a [`InternLMModel`]. It is used to instantiate
|
33 |
-
|
34 |
-
|
35 |
|
36 |
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
|
37 |
documentation from [`PretrainedConfig`] for more information.
|
@@ -80,7 +81,7 @@ class InternLMConfig(PretrainedConfig):
|
|
80 |
model_type = "internlm"
|
81 |
_auto_class = "AutoConfig"
|
82 |
|
83 |
-
def __init__(
|
84 |
self,
|
85 |
vocab_size=103168,
|
86 |
hidden_size=4096,
|
@@ -97,7 +98,6 @@ class InternLMConfig(PretrainedConfig):
|
|
97 |
eos_token_id=2,
|
98 |
tie_word_embeddings=False,
|
99 |
bias=True,
|
100 |
-
rotary={"base": 10000, "type": "dynamic"}, # pylint: disable=W0102
|
101 |
**kwargs,
|
102 |
):
|
103 |
self.vocab_size = vocab_size
|
@@ -111,11 +111,10 @@ class InternLMConfig(PretrainedConfig):
|
|
111 |
self.rms_norm_eps = rms_norm_eps
|
112 |
self.use_cache = use_cache
|
113 |
self.bias = bias
|
114 |
-
self.rotary = rotary
|
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 |
-
)
|
|
|
19 |
# limitations under the License.
|
20 |
""" InternLM model configuration"""
|
21 |
|
|
|
22 |
from transformers.utils import logging
|
23 |
+
from transformers.configuration_utils import PretrainedConfig
|
24 |
+
|
25 |
|
26 |
logger = logging.get_logger(__name__)
|
27 |
|
|
|
30 |
|
31 |
class InternLMConfig(PretrainedConfig):
|
32 |
r"""
|
33 |
+
This is the configuration class to store the configuration of a [`InternLMModel`]. It is used to instantiate an InternLM
|
34 |
+
model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
|
35 |
+
defaults will yield a similar configuration to that of the InternLM-7B.
|
36 |
|
37 |
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
|
38 |
documentation from [`PretrainedConfig`] for more information.
|
|
|
81 |
model_type = "internlm"
|
82 |
_auto_class = "AutoConfig"
|
83 |
|
84 |
+
def __init__(
|
85 |
self,
|
86 |
vocab_size=103168,
|
87 |
hidden_size=4096,
|
|
|
98 |
eos_token_id=2,
|
99 |
tie_word_embeddings=False,
|
100 |
bias=True,
|
|
|
101 |
**kwargs,
|
102 |
):
|
103 |
self.vocab_size = vocab_size
|
|
|
111 |
self.rms_norm_eps = rms_norm_eps
|
112 |
self.use_cache = use_cache
|
113 |
self.bias = bias
|
|
|
114 |
super().__init__(
|
115 |
pad_token_id=pad_token_id,
|
116 |
bos_token_id=bos_token_id,
|
117 |
eos_token_id=eos_token_id,
|
118 |
tie_word_embeddings=tie_word_embeddings,
|
119 |
**kwargs,
|
120 |
+
)
|
modeling_internlm.py
CHANGED
@@ -19,36 +19,26 @@
|
|
19 |
# limitations under the License.
|
20 |
""" PyTorch InternLM model."""
|
21 |
import math
|
22 |
-
import queue
|
23 |
-
import threading
|
24 |
from typing import List, Optional, Tuple, Union
|
|
|
25 |
|
26 |
import torch
|
27 |
import torch.utils.checkpoint
|
28 |
from torch import nn
|
29 |
from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
|
|
|
30 |
from transformers.activations import ACT2FN
|
31 |
-
from transformers.
|
32 |
-
from transformers.modeling_outputs import (
|
33 |
-
BaseModelOutputWithPast,
|
34 |
-
CausalLMOutputWithPast,
|
35 |
-
SequenceClassifierOutputWithPast,
|
36 |
-
)
|
37 |
from transformers.modeling_utils import PreTrainedModel
|
38 |
-
from transformers.
|
39 |
-
|
40 |
-
add_start_docstrings_to_model_forward,
|
41 |
-
logging,
|
42 |
-
replace_return_docstrings,
|
43 |
-
)
|
44 |
-
|
45 |
from .configuration_internlm import InternLMConfig
|
46 |
|
|
|
47 |
logger = logging.get_logger(__name__)
|
48 |
|
49 |
_CONFIG_FOR_DOC = "InternLMConfig"
|
50 |
|
51 |
-
|
52 |
# Copied from transformers.models.bart.modeling_bart._make_causal_mask
|
53 |
def _make_causal_mask(
|
54 |
input_ids_shape: torch.Size, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0
|
@@ -83,8 +73,6 @@ def _expand_mask(mask: torch.Tensor, dtype: torch.dtype, tgt_len: Optional[int]
|
|
83 |
|
84 |
|
85 |
class InternLMRMSNorm(nn.Module):
|
86 |
-
"""RMSNorm implemention."""
|
87 |
-
|
88 |
def __init__(self, hidden_size, eps=1e-6):
|
89 |
"""
|
90 |
InternLMRMSNorm is equivalent to T5LayerNorm
|
@@ -105,14 +93,6 @@ class InternLMRMSNorm(nn.Module):
|
|
105 |
|
106 |
|
107 |
class InternLMRotaryEmbedding(torch.nn.Module):
|
108 |
-
"""Implement InternLM's rotary embedding.
|
109 |
-
|
110 |
-
Args:
|
111 |
-
dim (int): Characteristic dimension of each self-attentional head.
|
112 |
-
max_position_embeddings (int, optional): Model's training length. Defaults to 2048.
|
113 |
-
base (int, optional): The rotation position encodes the rotation Angle base number. Defaults to 10000.
|
114 |
-
device (Any, optional): Running device. Defaults to None.
|
115 |
-
"""
|
116 |
def __init__(self, dim, max_position_embeddings=2048, base=10000, device=None):
|
117 |
super().__init__()
|
118 |
inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float().to(device) / dim))
|
@@ -144,66 +124,6 @@ class InternLMRotaryEmbedding(torch.nn.Module):
|
|
144 |
)
|
145 |
|
146 |
|
147 |
-
class InternLMDynamicNTKScalingRotaryEmbedding(torch.nn.Module):
|
148 |
-
"""Implement InternLM's DyanmicNTK extrapolation method, thereby broadening the model support context to 16K.
|
149 |
-
|
150 |
-
Args:
|
151 |
-
dim (int): Characteristic dimension of each self-attentional head.
|
152 |
-
max_position_embeddings (int, optional): Model's training length. Defaults to 2048.
|
153 |
-
base (int, optional): The rotation position encodes the rotation Angle base number. Defaults to 10000.
|
154 |
-
device (Any, optional): Running device. Defaults to None.
|
155 |
-
scaling_factor (float, optional): NTK method extrapolation coefficient. Defaults to 1.0.
|
156 |
-
"""
|
157 |
-
|
158 |
-
def __init__(self, dim, max_position_embeddings=2048, base=10000, device=None, scaling_factor=1.0):
|
159 |
-
super().__init__()
|
160 |
-
inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float().to(device) / dim))
|
161 |
-
self.register_buffer("inv_freq", inv_freq)
|
162 |
-
self.dim = dim
|
163 |
-
self.base = base
|
164 |
-
self.scaling_factor = scaling_factor
|
165 |
-
|
166 |
-
# Build here to make `torch.jit.trace` work.
|
167 |
-
self.max_position_embeddings = max_position_embeddings
|
168 |
-
self.max_seq_len_cached = max_position_embeddings
|
169 |
-
t = torch.arange(self.max_seq_len_cached, device=self.inv_freq.device, dtype=self.inv_freq.dtype)
|
170 |
-
freqs = torch.einsum("i,j->ij", t, self.inv_freq)
|
171 |
-
# Different from paper, but it uses a different permutation in order to obtain the same calculation
|
172 |
-
emb = torch.cat((freqs, freqs), dim=-1)
|
173 |
-
self.register_buffer("cos_cached", emb.cos()[None, None, :, :], persistent=False)
|
174 |
-
self.register_buffer("sin_cached", emb.sin()[None, None, :, :], persistent=False)
|
175 |
-
|
176 |
-
def _update_cached(self, x, seq_len=None):
|
177 |
-
self.max_seq_len_cached = max(seq_len, self.max_position_embeddings)
|
178 |
-
if seq_len > self.max_position_embeddings:
|
179 |
-
base = self.base * (
|
180 |
-
(self.scaling_factor * seq_len / self.max_position_embeddings) - (self.scaling_factor - 1)
|
181 |
-
) ** (self.dim / (self.dim - 2))
|
182 |
-
inv_freq = 1.0 / (base ** (torch.arange(0, self.dim, 2).float().to(x.device) / self.dim))
|
183 |
-
else:
|
184 |
-
inv_freq = self.inv_freq
|
185 |
-
t = torch.arange(self.max_seq_len_cached, device=inv_freq.device, dtype=inv_freq.dtype)
|
186 |
-
freqs = torch.einsum("i,j->ij", t, inv_freq)
|
187 |
-
emb = torch.cat((freqs, freqs), dim=-1)
|
188 |
-
self.register_buffer("cos_cached", emb.cos()[None, None, :, :], persistent=False)
|
189 |
-
self.register_buffer("sin_cached", emb.sin()[None, None, :, :], persistent=False)
|
190 |
-
|
191 |
-
def forward(self, x, seq_len=None):
|
192 |
-
# x: [bs, num_attention_heads, seq_len, head_size]
|
193 |
-
# This `if` block is unlikely to be run after we build sin/cos in `__init__`. Keep the logic here just in case.
|
194 |
-
if seq_len <= self.max_position_embeddings:
|
195 |
-
# Reset the tables if the sequence length has changed,
|
196 |
-
if self.max_seq_len_cached > self.max_position_embeddings:
|
197 |
-
self._update_cached(x, seq_len)
|
198 |
-
else:
|
199 |
-
self._update_cached(x, seq_len)
|
200 |
-
|
201 |
-
return (
|
202 |
-
self.cos_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
|
203 |
-
self.sin_cached[:, :, :seq_len, ...].to(dtype=x.dtype),
|
204 |
-
)
|
205 |
-
|
206 |
-
|
207 |
def rotate_half(x):
|
208 |
"""Rotates half the hidden dims of the input."""
|
209 |
x1 = x[..., : x.shape[-1] // 2]
|
@@ -215,18 +135,10 @@ def apply_rotary_pos_emb(q, k, cos, sin, position_ids):
|
|
215 |
# The first two dimensions of cos and sin are always 1, so we can `squeeze` them.
|
216 |
cos = cos.squeeze(1).squeeze(0) # [seq_len, dim]
|
217 |
sin = sin.squeeze(1).squeeze(0) # [seq_len, dim]
|
218 |
-
cos = cos.unsqueeze(
|
219 |
-
sin = sin.unsqueeze(
|
220 |
-
|
221 |
-
|
222 |
-
else:
|
223 |
-
q_embed = (q * cos) + (rotate_half(q) * sin)
|
224 |
-
|
225 |
-
if k.size(2) == 1:
|
226 |
-
k_embed = (k * cos[:, :, -1, :]) + (rotate_half(k) * sin[:, :, -1, :])
|
227 |
-
else:
|
228 |
-
k_embed = (k * cos) + (rotate_half(k) * sin)
|
229 |
-
|
230 |
return q_embed, k_embed
|
231 |
|
232 |
|
@@ -267,25 +179,7 @@ class InternLMAttention(nn.Module):
|
|
267 |
self.k_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.bias)
|
268 |
self.v_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.bias)
|
269 |
self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=config.bias)
|
270 |
-
self.rotary_emb = self.
|
271 |
-
|
272 |
-
def _init_rope(self):
|
273 |
-
if self.config.rotary["type"] == "origin":
|
274 |
-
self.rotary_emb = InternLMRotaryEmbedding(
|
275 |
-
self.head_dim,
|
276 |
-
max_position_embeddings=self.max_position_embeddings,
|
277 |
-
base=self.config.rotary["base"],
|
278 |
-
)
|
279 |
-
elif self.config.rotary["type"] == "dynamic":
|
280 |
-
self.rotary_emb = InternLMDynamicNTKScalingRotaryEmbedding(
|
281 |
-
self.head_dim,
|
282 |
-
max_position_embeddings=self.max_position_embeddings,
|
283 |
-
base=self.config.rotary["base"],
|
284 |
-
scaling_factor=self.config.rotary.get("scaling_factor", 1.0),
|
285 |
-
)
|
286 |
-
else:
|
287 |
-
raise ValueError("Currently we only support rotary embedding's type being one of ('origin', 'dynamic').")
|
288 |
-
return self.rotary_emb
|
289 |
|
290 |
def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int):
|
291 |
return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous()
|
@@ -305,18 +199,20 @@ class InternLMAttention(nn.Module):
|
|
305 |
key_states = self.k_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
|
306 |
value_states = self.v_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
|
307 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
308 |
if past_key_value is not None:
|
309 |
# reuse k, v, self_attention
|
310 |
key_states = torch.cat([past_key_value[0], key_states], dim=2)
|
311 |
value_states = torch.cat([past_key_value[1], value_states], dim=2)
|
312 |
|
313 |
-
# print(use_cache)
|
314 |
past_key_value = (key_states, value_states) if use_cache else None
|
315 |
|
316 |
-
kv_seq_len = key_states.shape[-2]
|
317 |
-
cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
|
318 |
-
query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, position_ids)
|
319 |
-
|
320 |
attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
|
321 |
|
322 |
if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
|
@@ -426,9 +322,11 @@ INTERNLM_START_DOCSTRING = r"""
|
|
426 |
This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
|
427 |
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
|
428 |
etc.)
|
|
|
429 |
This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
|
430 |
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
|
431 |
and behavior.
|
|
|
432 |
Parameters:
|
433 |
config ([`InternLMConfig`]):
|
434 |
Model configuration class with all the parameters of the model. Initializing with a config file does not
|
@@ -469,34 +367,44 @@ INTERNLM_INPUTS_DOCSTRING = r"""
|
|
469 |
input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
|
470 |
Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
|
471 |
it.
|
|
|
472 |
Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
|
473 |
[`PreTrainedTokenizer.__call__`] for details.
|
|
|
474 |
[What are input IDs?](../glossary#input-ids)
|
475 |
attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
|
476 |
Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
|
|
|
477 |
- 1 for tokens that are **not masked**,
|
478 |
- 0 for tokens that are **masked**.
|
|
|
479 |
[What are attention masks?](../glossary#attention-mask)
|
|
|
480 |
Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
|
481 |
[`PreTrainedTokenizer.__call__`] for details.
|
|
|
482 |
If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
|
483 |
`past_key_values`).
|
|
|
484 |
If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
|
485 |
and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
|
486 |
information on the default strategy.
|
|
|
487 |
- 1 indicates the head is **not masked**,
|
488 |
- 0 indicates the head is **masked**.
|
489 |
position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
|
490 |
Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
|
491 |
config.n_positions - 1]`.
|
|
|
492 |
[What are position IDs?](../glossary#position-ids)
|
493 |
-
past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or
|
494 |
-
when `config.use_cache=True`):
|
495 |
Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
|
496 |
`(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
|
497 |
`(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
|
|
|
498 |
Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
|
499 |
blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.
|
|
|
500 |
If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
|
501 |
don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
|
502 |
`decoder_input_ids` of shape `(batch_size, sequence_length)`.
|
@@ -525,16 +433,13 @@ INTERNLM_INPUTS_DOCSTRING = r"""
|
|
525 |
class InternLMModel(InternLMPreTrainedModel):
|
526 |
"""
|
527 |
Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`InternLMDecoderLayer`]
|
|
|
528 |
Args:
|
529 |
config: InternLMConfig
|
530 |
"""
|
531 |
-
|
532 |
_auto_class = "AutoModel"
|
533 |
|
534 |
def __init__(self, config: InternLMConfig):
|
535 |
-
assert (0), 'Inference by transformers is currently not supported, ' \
|
536 |
-
'please follow README to convert the model ' \
|
537 |
-
'and use lmdeploy (https://github.com/InternLM/lmdeploy) for inference.'
|
538 |
super().__init__(config)
|
539 |
self.padding_idx = config.pad_token_id
|
540 |
self.vocab_size = config.vocab_size
|
@@ -757,14 +662,20 @@ class InternLMForCausalLM(InternLMPreTrainedModel):
|
|
757 |
Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
|
758 |
config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
|
759 |
(masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
|
|
|
760 |
Returns:
|
|
|
761 |
Example:
|
|
|
762 |
```python
|
763 |
>>> from transformers import AutoTokenizer, InternLMForCausalLM
|
|
|
764 |
>>> model = InternLMForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
|
765 |
>>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
|
|
|
766 |
>>> prompt = "Hey, are you consciours? Can you talk to me?"
|
767 |
>>> inputs = tokenizer(prompt, return_tensors="pt")
|
|
|
768 |
>>> # Generate
|
769 |
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
|
770 |
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
|
@@ -854,56 +765,52 @@ class InternLMForCausalLM(InternLMPreTrainedModel):
|
|
854 |
for layer_past in past_key_values:
|
855 |
reordered_past += (tuple(past_state.index_select(0, beam_idx) for past_state in layer_past),)
|
856 |
return reordered_past
|
857 |
-
|
858 |
def build_inputs(self, tokenizer, query: str, history: List[Tuple[str, str]] = []):
|
859 |
prompt = ""
|
860 |
for record in history:
|
861 |
-
prompt += f"""
|
|
|
|
|
862 |
prompt += f"""<|User|>:{query}<eoh>\n<|Bot|>:"""
|
863 |
return tokenizer([prompt], return_tensors="pt")
|
864 |
-
|
865 |
@torch.no_grad()
|
866 |
-
def chat(
|
867 |
-
|
868 |
-
|
869 |
-
|
870 |
-
|
871 |
-
|
872 |
-
|
873 |
-
|
874 |
-
|
875 |
-
|
876 |
-
**kwargs,
|
877 |
-
):
|
878 |
inputs = self.build_inputs(tokenizer, query, history)
|
879 |
inputs = {k: v.to(self.device) for k, v in inputs.items() if torch.is_tensor(v)}
|
880 |
-
outputs = self.generate(
|
881 |
-
|
882 |
-
|
883 |
-
|
884 |
-
|
885 |
-
|
886 |
-
|
887 |
-
|
888 |
-
)
|
889 |
-
outputs = outputs[0].cpu().tolist()[len(inputs["input_ids"][0]) :]
|
890 |
response = tokenizer.decode(outputs, skip_special_tokens=True)
|
891 |
response = response.split("<eoa>")[0]
|
892 |
history = history + [(query, response)]
|
893 |
return response, history
|
894 |
-
|
895 |
@torch.no_grad()
|
896 |
-
def stream_chat(
|
897 |
-
|
898 |
-
|
899 |
-
|
900 |
-
|
901 |
-
|
902 |
-
|
903 |
-
|
904 |
-
|
905 |
-
**kwargs,
|
906 |
-
):
|
907 |
"""
|
908 |
Return a generator in format: (response, history)
|
909 |
Eg.
|
@@ -949,12 +856,12 @@ class InternLMForCausalLM(InternLMPreTrainedModel):
|
|
949 |
tokenizer=tokenizer,
|
950 |
query=query,
|
951 |
streamer=ChatStreamer(tokenizer=tokenizer),
|
952 |
-
history=history,
|
953 |
max_new_tokens=max_new_tokens,
|
954 |
do_sample=do_sample,
|
955 |
temperature=temperature,
|
956 |
top_p=top_p,
|
957 |
-
**kwargs
|
958 |
)
|
959 |
|
960 |
def consumer():
|
@@ -972,8 +879,10 @@ class InternLMForCausalLM(InternLMPreTrainedModel):
|
|
972 |
@add_start_docstrings(
|
973 |
"""
|
974 |
The InternLM Model transformer with a sequence classification head on top (linear layer).
|
|
|
975 |
[`InternLMForSequenceClassification`] uses the last token in order to do the classification, as other causal models
|
976 |
(e.g. GPT-2) do.
|
|
|
977 |
Since it does classification on the last token, it requires to know the position of the last token. If a
|
978 |
`pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If
|
979 |
no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the
|
|
|
19 |
# limitations under the License.
|
20 |
""" PyTorch InternLM model."""
|
21 |
import math
|
|
|
|
|
22 |
from typing import List, Optional, Tuple, Union
|
23 |
+
import threading, queue
|
24 |
|
25 |
import torch
|
26 |
import torch.utils.checkpoint
|
27 |
from torch import nn
|
28 |
from torch.nn import BCEWithLogitsLoss, CrossEntropyLoss, MSELoss
|
29 |
+
|
30 |
from transformers.activations import ACT2FN
|
31 |
+
from transformers.modeling_outputs import BaseModelOutputWithPast, CausalLMOutputWithPast, SequenceClassifierOutputWithPast
|
|
|
|
|
|
|
|
|
|
|
32 |
from transformers.modeling_utils import PreTrainedModel
|
33 |
+
from transformers.generation.streamers import BaseStreamer
|
34 |
+
from transformers.utils import add_start_docstrings, add_start_docstrings_to_model_forward, logging, replace_return_docstrings
|
|
|
|
|
|
|
|
|
|
|
35 |
from .configuration_internlm import InternLMConfig
|
36 |
|
37 |
+
|
38 |
logger = logging.get_logger(__name__)
|
39 |
|
40 |
_CONFIG_FOR_DOC = "InternLMConfig"
|
41 |
|
|
|
42 |
# Copied from transformers.models.bart.modeling_bart._make_causal_mask
|
43 |
def _make_causal_mask(
|
44 |
input_ids_shape: torch.Size, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0
|
|
|
73 |
|
74 |
|
75 |
class InternLMRMSNorm(nn.Module):
|
|
|
|
|
76 |
def __init__(self, hidden_size, eps=1e-6):
|
77 |
"""
|
78 |
InternLMRMSNorm is equivalent to T5LayerNorm
|
|
|
93 |
|
94 |
|
95 |
class InternLMRotaryEmbedding(torch.nn.Module):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
def __init__(self, dim, max_position_embeddings=2048, base=10000, device=None):
|
97 |
super().__init__()
|
98 |
inv_freq = 1.0 / (base ** (torch.arange(0, dim, 2).float().to(device) / dim))
|
|
|
124 |
)
|
125 |
|
126 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
127 |
def rotate_half(x):
|
128 |
"""Rotates half the hidden dims of the input."""
|
129 |
x1 = x[..., : x.shape[-1] // 2]
|
|
|
135 |
# The first two dimensions of cos and sin are always 1, so we can `squeeze` them.
|
136 |
cos = cos.squeeze(1).squeeze(0) # [seq_len, dim]
|
137 |
sin = sin.squeeze(1).squeeze(0) # [seq_len, dim]
|
138 |
+
cos = cos[position_ids].unsqueeze(1) # [bs, 1, seq_len, dim]
|
139 |
+
sin = sin[position_ids].unsqueeze(1) # [bs, 1, seq_len, dim]
|
140 |
+
q_embed = (q * cos) + (rotate_half(q) * sin)
|
141 |
+
k_embed = (k * cos) + (rotate_half(k) * sin)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
142 |
return q_embed, k_embed
|
143 |
|
144 |
|
|
|
179 |
self.k_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.bias)
|
180 |
self.v_proj = nn.Linear(self.hidden_size, self.num_heads * self.head_dim, bias=config.bias)
|
181 |
self.o_proj = nn.Linear(self.num_heads * self.head_dim, self.hidden_size, bias=config.bias)
|
182 |
+
self.rotary_emb = InternLMRotaryEmbedding(self.head_dim, max_position_embeddings=self.max_position_embeddings)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
183 |
|
184 |
def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int):
|
185 |
return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous()
|
|
|
199 |
key_states = self.k_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
|
200 |
value_states = self.v_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2)
|
201 |
|
202 |
+
kv_seq_len = key_states.shape[-2]
|
203 |
+
if past_key_value is not None:
|
204 |
+
kv_seq_len += past_key_value[0].shape[-2]
|
205 |
+
cos, sin = self.rotary_emb(value_states, seq_len=kv_seq_len)
|
206 |
+
query_states, key_states = apply_rotary_pos_emb(query_states, key_states, cos, sin, position_ids)
|
207 |
+
# [bsz, nh, t, hd]
|
208 |
+
|
209 |
if past_key_value is not None:
|
210 |
# reuse k, v, self_attention
|
211 |
key_states = torch.cat([past_key_value[0], key_states], dim=2)
|
212 |
value_states = torch.cat([past_key_value[1], value_states], dim=2)
|
213 |
|
|
|
214 |
past_key_value = (key_states, value_states) if use_cache else None
|
215 |
|
|
|
|
|
|
|
|
|
216 |
attn_weights = torch.matmul(query_states, key_states.transpose(2, 3)) / math.sqrt(self.head_dim)
|
217 |
|
218 |
if attn_weights.size() != (bsz, self.num_heads, q_len, kv_seq_len):
|
|
|
322 |
This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
|
323 |
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
|
324 |
etc.)
|
325 |
+
|
326 |
This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
|
327 |
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
|
328 |
and behavior.
|
329 |
+
|
330 |
Parameters:
|
331 |
config ([`InternLMConfig`]):
|
332 |
Model configuration class with all the parameters of the model. Initializing with a config file does not
|
|
|
367 |
input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
|
368 |
Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
|
369 |
it.
|
370 |
+
|
371 |
Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
|
372 |
[`PreTrainedTokenizer.__call__`] for details.
|
373 |
+
|
374 |
[What are input IDs?](../glossary#input-ids)
|
375 |
attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
|
376 |
Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:
|
377 |
+
|
378 |
- 1 for tokens that are **not masked**,
|
379 |
- 0 for tokens that are **masked**.
|
380 |
+
|
381 |
[What are attention masks?](../glossary#attention-mask)
|
382 |
+
|
383 |
Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
|
384 |
[`PreTrainedTokenizer.__call__`] for details.
|
385 |
+
|
386 |
If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
|
387 |
`past_key_values`).
|
388 |
+
|
389 |
If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
|
390 |
and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
|
391 |
information on the default strategy.
|
392 |
+
|
393 |
- 1 indicates the head is **not masked**,
|
394 |
- 0 indicates the head is **masked**.
|
395 |
position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
|
396 |
Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
|
397 |
config.n_positions - 1]`.
|
398 |
+
|
399 |
[What are position IDs?](../glossary#position-ids)
|
400 |
+
past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
|
|
|
401 |
Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
|
402 |
`(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
|
403 |
`(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.
|
404 |
+
|
405 |
Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
|
406 |
blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.
|
407 |
+
|
408 |
If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
|
409 |
don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
|
410 |
`decoder_input_ids` of shape `(batch_size, sequence_length)`.
|
|
|
433 |
class InternLMModel(InternLMPreTrainedModel):
|
434 |
"""
|
435 |
Transformer decoder consisting of *config.num_hidden_layers* layers. Each layer is a [`InternLMDecoderLayer`]
|
436 |
+
|
437 |
Args:
|
438 |
config: InternLMConfig
|
439 |
"""
|
|
|
440 |
_auto_class = "AutoModel"
|
441 |
|
442 |
def __init__(self, config: InternLMConfig):
|
|
|
|
|
|
|
443 |
super().__init__(config)
|
444 |
self.padding_idx = config.pad_token_id
|
445 |
self.vocab_size = config.vocab_size
|
|
|
662 |
Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
|
663 |
config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
|
664 |
(masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.
|
665 |
+
|
666 |
Returns:
|
667 |
+
|
668 |
Example:
|
669 |
+
|
670 |
```python
|
671 |
>>> from transformers import AutoTokenizer, InternLMForCausalLM
|
672 |
+
|
673 |
>>> model = InternLMForCausalLM.from_pretrained(PATH_TO_CONVERTED_WEIGHTS)
|
674 |
>>> tokenizer = AutoTokenizer.from_pretrained(PATH_TO_CONVERTED_TOKENIZER)
|
675 |
+
|
676 |
>>> prompt = "Hey, are you consciours? Can you talk to me?"
|
677 |
>>> inputs = tokenizer(prompt, return_tensors="pt")
|
678 |
+
|
679 |
>>> # Generate
|
680 |
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
|
681 |
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
|
|
|
765 |
for layer_past in past_key_values:
|
766 |
reordered_past += (tuple(past_state.index_select(0, beam_idx) for past_state in layer_past),)
|
767 |
return reordered_past
|
768 |
+
|
769 |
def build_inputs(self, tokenizer, query: str, history: List[Tuple[str, str]] = []):
|
770 |
prompt = ""
|
771 |
for record in history:
|
772 |
+
prompt += f"""<s><|User|>:{record[0]}<eoh>\n<|Bot|>:{record[1]}<eoa>\n"""
|
773 |
+
if len(prompt) == 0:
|
774 |
+
prompt += "<s>"
|
775 |
prompt += f"""<|User|>:{query}<eoh>\n<|Bot|>:"""
|
776 |
return tokenizer([prompt], return_tensors="pt")
|
777 |
+
|
778 |
@torch.no_grad()
|
779 |
+
def chat(self,
|
780 |
+
tokenizer,
|
781 |
+
query: str,
|
782 |
+
history: List[Tuple[str, str]] = [],
|
783 |
+
streamer: Optional[BaseStreamer] = None,
|
784 |
+
max_new_tokens: int = 1024,
|
785 |
+
do_sample: bool = True,
|
786 |
+
temperature: float = 0.8,
|
787 |
+
top_p: float = 0.8,
|
788 |
+
**kwargs):
|
|
|
|
|
789 |
inputs = self.build_inputs(tokenizer, query, history)
|
790 |
inputs = {k: v.to(self.device) for k, v in inputs.items() if torch.is_tensor(v)}
|
791 |
+
outputs = self.generate(**inputs,
|
792 |
+
streamer=streamer,
|
793 |
+
max_new_tokens=max_new_tokens,
|
794 |
+
do_sample=do_sample,
|
795 |
+
temperature=temperature,
|
796 |
+
top_p=top_p,
|
797 |
+
**kwargs)
|
798 |
+
outputs = outputs[0].cpu().tolist()[len(inputs["input_ids"][0]):]
|
|
|
|
|
799 |
response = tokenizer.decode(outputs, skip_special_tokens=True)
|
800 |
response = response.split("<eoa>")[0]
|
801 |
history = history + [(query, response)]
|
802 |
return response, history
|
803 |
+
|
804 |
@torch.no_grad()
|
805 |
+
def stream_chat(self,
|
806 |
+
tokenizer,
|
807 |
+
query: str,
|
808 |
+
history: List[Tuple[str, str]] = [],
|
809 |
+
max_new_tokens: int = 1024,
|
810 |
+
do_sample: bool = True,
|
811 |
+
temperature: float = 0.8,
|
812 |
+
top_p: float = 0.8,
|
813 |
+
**kwargs):
|
|
|
|
|
814 |
"""
|
815 |
Return a generator in format: (response, history)
|
816 |
Eg.
|
|
|
856 |
tokenizer=tokenizer,
|
857 |
query=query,
|
858 |
streamer=ChatStreamer(tokenizer=tokenizer),
|
859 |
+
history=history,
|
860 |
max_new_tokens=max_new_tokens,
|
861 |
do_sample=do_sample,
|
862 |
temperature=temperature,
|
863 |
top_p=top_p,
|
864 |
+
**kwargs
|
865 |
)
|
866 |
|
867 |
def consumer():
|
|
|
879 |
@add_start_docstrings(
|
880 |
"""
|
881 |
The InternLM Model transformer with a sequence classification head on top (linear layer).
|
882 |
+
|
883 |
[`InternLMForSequenceClassification`] uses the last token in order to do the classification, as other causal models
|
884 |
(e.g. GPT-2) do.
|
885 |
+
|
886 |
Since it does classification on the last token, it requires to know the position of the last token. If a
|
887 |
`pad_token_id` is defined in the configuration, it finds the last token that is not a padding token in each row. If
|
888 |
no `pad_token_id` is defined, it simply takes the last value in each row of the batch. Since it cannot guess the
|