tomer-deci
commited on
Commit
•
49422ae
1
Parent(s):
64cf789
added support for text-generation pipeline
Browse files- config.json +2 -1
- modeling_decilm.py +9 -10
- version_check.py +3 -3
config.json
CHANGED
@@ -25,5 +25,6 @@
|
|
25 |
"use_bfloat16": true,
|
26 |
"transformers_version": "4.35.2",
|
27 |
"use_cache": true,
|
28 |
-
"vocab_size": 32000
|
|
|
29 |
}
|
|
|
25 |
"use_bfloat16": true,
|
26 |
"transformers_version": "4.35.2",
|
27 |
"use_cache": true,
|
28 |
+
"vocab_size": 32000,
|
29 |
+
"tokenizer_class": "LlamaTokenizer"
|
30 |
}
|
modeling_decilm.py
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
# coding=utf-8
|
2 |
-
# Copyright and license
|
3 |
""" PyTorch DeciLM model."""
|
4 |
from .version_check import check_transformers_version
|
5 |
|
@@ -11,18 +11,18 @@ import torch
|
|
11 |
import torch.nn.functional as F
|
12 |
import torch.utils.checkpoint
|
13 |
from torch import nn
|
|
|
|
|
14 |
|
|
|
|
|
15 |
from .transformers_v4_35_2__modeling_llama import LlamaMLP, LlamaRMSNorm, LlamaAttention, apply_rotary_pos_emb, \
|
16 |
repeat_kv, LlamaPreTrainedModel, LLAMA_START_DOCSTRING, LlamaDecoderLayer, LlamaForCausalLM, LlamaModel, \
|
17 |
BaseModelOutputWithPast, LLAMA_INPUTS_DOCSTRING
|
18 |
-
from .transformers_v4_35_2__modeling_attn_mask_utils import _prepare_4d_causal_attention_mask
|
19 |
-
from transformers.utils import add_start_docstrings, add_start_docstrings_to_model_forward, logging
|
20 |
-
|
21 |
-
from .configuration_decilm import DeciLMConfig
|
22 |
-
|
23 |
-
logger = logging.get_logger(__name__)
|
24 |
|
|
|
25 |
_CONFIG_FOR_DOC = "DeciLMConfig"
|
|
|
26 |
|
27 |
|
28 |
class DeciLMAttention(LlamaAttention):
|
@@ -61,7 +61,7 @@ class DeciLMAttention(LlamaAttention):
|
|
61 |
past_key_value: Optional[Tuple[torch.Tensor]] = None,
|
62 |
output_attentions: bool = False,
|
63 |
use_cache: bool = False,
|
64 |
-
|
65 |
) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
|
66 |
bsz, q_len, _ = hidden_states.size()
|
67 |
is_decode = past_key_value is not None
|
@@ -135,8 +135,7 @@ class DeciLMAttention(LlamaAttention):
|
|
135 |
else:
|
136 |
attn_output = self.o_proj(attn_output)
|
137 |
|
138 |
-
|
139 |
-
attn_weights = None
|
140 |
|
141 |
return attn_output, attn_weights, past_key_value
|
142 |
|
|
|
1 |
# coding=utf-8
|
2 |
+
# Copyright and license in the repo.
|
3 |
""" PyTorch DeciLM model."""
|
4 |
from .version_check import check_transformers_version
|
5 |
|
|
|
11 |
import torch.nn.functional as F
|
12 |
import torch.utils.checkpoint
|
13 |
from torch import nn
|
14 |
+
from transformers.models.auto.modeling_auto import MODEL_FOR_CAUSAL_LM_MAPPING_NAMES
|
15 |
+
from transformers.utils import add_start_docstrings, add_start_docstrings_to_model_forward, logging
|
16 |
|
17 |
+
from .configuration_decilm import DeciLMConfig
|
18 |
+
from .transformers_v4_35_2__modeling_attn_mask_utils import _prepare_4d_causal_attention_mask
|
19 |
from .transformers_v4_35_2__modeling_llama import LlamaMLP, LlamaRMSNorm, LlamaAttention, apply_rotary_pos_emb, \
|
20 |
repeat_kv, LlamaPreTrainedModel, LLAMA_START_DOCSTRING, LlamaDecoderLayer, LlamaForCausalLM, LlamaModel, \
|
21 |
BaseModelOutputWithPast, LLAMA_INPUTS_DOCSTRING
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
|
23 |
+
MODEL_FOR_CAUSAL_LM_MAPPING_NAMES["deci"] = "DeciLMForCausalLM"
|
24 |
_CONFIG_FOR_DOC = "DeciLMConfig"
|
25 |
+
logger = logging.get_logger(__name__)
|
26 |
|
27 |
|
28 |
class DeciLMAttention(LlamaAttention):
|
|
|
61 |
past_key_value: Optional[Tuple[torch.Tensor]] = None,
|
62 |
output_attentions: bool = False,
|
63 |
use_cache: bool = False,
|
64 |
+
**kwargs,
|
65 |
) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]:
|
66 |
bsz, q_len, _ = hidden_states.size()
|
67 |
is_decode = past_key_value is not None
|
|
|
135 |
else:
|
136 |
attn_output = self.o_proj(attn_output)
|
137 |
|
138 |
+
attn_weights = None
|
|
|
139 |
|
140 |
return attn_output, attn_weights, past_key_value
|
141 |
|
version_check.py
CHANGED
@@ -1,11 +1,11 @@
|
|
1 |
import transformers
|
2 |
from packaging import version
|
3 |
|
4 |
-
|
5 |
|
6 |
|
7 |
def check_transformers_version():
|
8 |
-
if version.parse(transformers.__version__) < version.parse(
|
9 |
raise ImportError(
|
10 |
-
f"You are using transformers=={transformers.__version__}, but transformers>={
|
11 |
)
|
|
|
1 |
import transformers
|
2 |
from packaging import version
|
3 |
|
4 |
+
MIN_VERSION = "4.35.2"
|
5 |
|
6 |
|
7 |
def check_transformers_version():
|
8 |
+
if version.parse(transformers.__version__) < version.parse(MIN_VERSION):
|
9 |
raise ImportError(
|
10 |
+
f"You are using transformers=={transformers.__version__}, but transformers>={MIN_VERSION} is required to use DeciLM. Please upgrade transformers."
|
11 |
)
|