refactor neft patch to be more re-usable similar to trl's impl (#796)
Browse files
gitbook/README.md
CHANGED
@@ -1,2 +1 @@
|
|
1 |
# Page
|
2 |
-
|
|
|
1 |
# Page
|
|
src/axolotl/monkeypatch/llama_embeddings_hijack.py
DELETED
@@ -1,40 +0,0 @@
|
|
1 |
-
"""
|
2 |
-
patch to add noisy embeddings per https://arxiv.org/abs/2310.05914
|
3 |
-
"""
|
4 |
-
|
5 |
-
import torch
|
6 |
-
import transformers.models.llama.modeling_llama
|
7 |
-
from transformers.utils import logging
|
8 |
-
|
9 |
-
logger = logging.get_logger(__name__)
|
10 |
-
|
11 |
-
|
12 |
-
def replace_llama_embeddings_with_uniform_distribution(noise_alpha=5):
|
13 |
-
# pylint: disable=duplicate-code
|
14 |
-
def noised_embed(orig_embed, noise_alpha, model):
|
15 |
-
def new_func(input_ids):
|
16 |
-
# during training, we add noise to the embedding
|
17 |
-
# during generation, we don't add noise to the embedding
|
18 |
-
if model.training:
|
19 |
-
embed_init = orig_embed(input_ids)
|
20 |
-
dims = torch.tensor(embed_init.size(1) * embed_init.size(2))
|
21 |
-
mag_norm = noise_alpha / torch.sqrt(dims)
|
22 |
-
return embed_init + torch.zeros_like(embed_init).uniform_(
|
23 |
-
-mag_norm, mag_norm
|
24 |
-
)
|
25 |
-
return orig_embed(input_ids)
|
26 |
-
|
27 |
-
return new_func
|
28 |
-
|
29 |
-
def post_init(orig_post_init):
|
30 |
-
def new_func(self):
|
31 |
-
orig_post_init(self)
|
32 |
-
self.embed_tokens.forward = noised_embed(
|
33 |
-
self.embed_tokens.forward, noise_alpha, self
|
34 |
-
)
|
35 |
-
|
36 |
-
return new_func
|
37 |
-
|
38 |
-
transformers.models.llama.modeling_llama.LlamaModel.post_init = post_init(
|
39 |
-
transformers.models.llama.modeling_llama.LlamaModel.post_init
|
40 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
src/axolotl/monkeypatch/mistral_embeddings_hijack.py
DELETED
@@ -1,40 +0,0 @@
|
|
1 |
-
"""
|
2 |
-
patch to add noisy embeddings per https://arxiv.org/abs/2310.05914
|
3 |
-
"""
|
4 |
-
|
5 |
-
import torch
|
6 |
-
import transformers.models.mistral.modeling_mistral
|
7 |
-
from transformers.utils import logging
|
8 |
-
|
9 |
-
logger = logging.get_logger(__name__)
|
10 |
-
|
11 |
-
|
12 |
-
def replace_mistral_embeddings_with_uniform_distribution(noise_alpha=5):
|
13 |
-
# pylint: disable=duplicate-code
|
14 |
-
def noised_embed(orig_embed, noise_alpha, model):
|
15 |
-
def new_func(input_ids):
|
16 |
-
# during training, we add noise to the embedding
|
17 |
-
# during generation, we don't add noise to the embedding
|
18 |
-
if model.training:
|
19 |
-
embed_init = orig_embed(input_ids)
|
20 |
-
dims = torch.tensor(embed_init.size(1) * embed_init.size(2))
|
21 |
-
mag_norm = noise_alpha / torch.sqrt(dims)
|
22 |
-
return embed_init + torch.zeros_like(embed_init).uniform_(
|
23 |
-
-mag_norm, mag_norm
|
24 |
-
)
|
25 |
-
return orig_embed(input_ids)
|
26 |
-
|
27 |
-
return new_func
|
28 |
-
|
29 |
-
def post_init(orig_post_init):
|
30 |
-
def new_func(self):
|
31 |
-
orig_post_init(self)
|
32 |
-
self.embed_tokens.forward = noised_embed(
|
33 |
-
self.embed_tokens.forward, noise_alpha, self
|
34 |
-
)
|
35 |
-
|
36 |
-
return new_func
|
37 |
-
|
38 |
-
transformers.models.mistral.modeling_mistral.MistralModel.post_init = post_init(
|
39 |
-
transformers.models.mistral.modeling_mistral.MistralModel.post_init
|
40 |
-
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
src/axolotl/monkeypatch/neft_embeddings.py
ADDED
@@ -0,0 +1,65 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
"""
|
2 |
+
patches implemented through the trainer hooks to enable NEFT/noisy embeddings per https://arxiv.org/abs/2310.05914
|
3 |
+
"""
|
4 |
+
import torch
|
5 |
+
from peft import PeftModel
|
6 |
+
from transformers import PreTrainedModel
|
7 |
+
|
8 |
+
|
9 |
+
def patch_neft(alpha, model):
|
10 |
+
embeddings = None
|
11 |
+
if isinstance(model, PreTrainedModel):
|
12 |
+
embeddings = model.get_input_embeddings()
|
13 |
+
if isinstance(model, PeftModel):
|
14 |
+
embeddings = model.base_model.get_input_embeddings()
|
15 |
+
if not embeddings:
|
16 |
+
raise ValueError(f"unhandled model class for neft: {model.__class__.__name__}")
|
17 |
+
embeddings.noisy_embedding_alpha = alpha
|
18 |
+
old_forward = embeddings.forward
|
19 |
+
|
20 |
+
# This hack seems to be needed to properly use a custom forward pass
|
21 |
+
# all credits to: https://discuss.pytorch.org/t/how-can-i-replace-the-forward-method-of-a-predefined-torchvision-model-with-my-customized-forward-function/54224/11
|
22 |
+
bound_method = neft_forward.__get__( # pylint: disable=no-value-for-parameter
|
23 |
+
embeddings, embeddings.__class__
|
24 |
+
)
|
25 |
+
setattr(embeddings, "forward", bound_method)
|
26 |
+
|
27 |
+
embeddings._old_forward = old_forward # pylint: disable=protected-access
|
28 |
+
return model
|
29 |
+
|
30 |
+
|
31 |
+
def unpatch_neft(model):
|
32 |
+
embeddings = None
|
33 |
+
if isinstance(model, PreTrainedModel):
|
34 |
+
embeddings = model.get_input_embeddings()
|
35 |
+
if isinstance(model, PeftModel):
|
36 |
+
embeddings = model.base_model.get_input_embeddings()
|
37 |
+
if not embeddings:
|
38 |
+
raise ValueError(f"unhandled model class for neft: {model.__class__.__name__}")
|
39 |
+
if hasattr(embeddings, "_old_forward"):
|
40 |
+
embeddings.forward = embeddings._old_forward # pylint: disable=protected-access
|
41 |
+
del embeddings._old_forward # pylint: disable=protected-access
|
42 |
+
del embeddings.noisy_embedding_alpha
|
43 |
+
|
44 |
+
|
45 |
+
def neft_forward(self, inputs: torch.Tensor):
|
46 |
+
embeddings = self._old_forward(inputs) # pylint: disable=protected-access
|
47 |
+
|
48 |
+
if self.training:
|
49 |
+
dims = torch.tensor(embeddings.size(1) * embeddings.size(2))
|
50 |
+
mag_norm = self.noisy_embedding_alpha / torch.sqrt(dims)
|
51 |
+
embeddings = embeddings + torch.zeros_like(embeddings).uniform_(
|
52 |
+
-mag_norm, mag_norm
|
53 |
+
)
|
54 |
+
|
55 |
+
return embeddings
|
56 |
+
|
57 |
+
|
58 |
+
def pretrain_hook(cfg, trainer):
|
59 |
+
if cfg.noisy_embedding_alpha:
|
60 |
+
trainer.model = patch_neft(cfg.noisy_embedding_alpha, trainer.model)
|
61 |
+
|
62 |
+
|
63 |
+
def post_train_hook(cfg, trainer):
|
64 |
+
if cfg.noisy_embedding_alpha:
|
65 |
+
unpatch_neft(trainer.model)
|
src/axolotl/train.py
CHANGED
@@ -16,6 +16,7 @@ from transformers.deepspeed import is_deepspeed_zero3_enabled
|
|
16 |
|
17 |
from axolotl.common.cli import TrainerCliArgs
|
18 |
from axolotl.logging_config import configure_logging
|
|
|
19 |
from axolotl.utils.dict import DictDefault
|
20 |
from axolotl.utils.models import load_model, load_tokenizer
|
21 |
from axolotl.utils.trainer import setup_trainer
|
@@ -107,6 +108,7 @@ def train(
|
|
107 |
if cfg.group_by_length:
|
108 |
LOG.info("hang tight... sorting dataset for group_by_length")
|
109 |
|
|
|
110 |
if cfg.flash_optimum:
|
111 |
with torch.backends.cuda.sdp_kernel(
|
112 |
enable_flash=True, enable_math=True, enable_mem_efficient=True
|
@@ -114,6 +116,7 @@ def train(
|
|
114 |
trainer.train(resume_from_checkpoint=resume_from_checkpoint)
|
115 |
else:
|
116 |
trainer.train(resume_from_checkpoint=resume_from_checkpoint)
|
|
|
117 |
|
118 |
LOG.info(f"Training Completed!!! Saving pre-trained model to {cfg.output_dir}")
|
119 |
|
@@ -163,3 +166,23 @@ def train(
|
|
163 |
trainer.create_model_card(model_name=cfg.output_dir.lstrip("./"))
|
164 |
|
165 |
return model, tokenizer
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
from axolotl.common.cli import TrainerCliArgs
|
18 |
from axolotl.logging_config import configure_logging
|
19 |
+
from axolotl.monkeypatch import neft_embeddings
|
20 |
from axolotl.utils.dict import DictDefault
|
21 |
from axolotl.utils.models import load_model, load_tokenizer
|
22 |
from axolotl.utils.trainer import setup_trainer
|
|
|
108 |
if cfg.group_by_length:
|
109 |
LOG.info("hang tight... sorting dataset for group_by_length")
|
110 |
|
111 |
+
pretrain_hooks(cfg, trainer)
|
112 |
if cfg.flash_optimum:
|
113 |
with torch.backends.cuda.sdp_kernel(
|
114 |
enable_flash=True, enable_math=True, enable_mem_efficient=True
|
|
|
116 |
trainer.train(resume_from_checkpoint=resume_from_checkpoint)
|
117 |
else:
|
118 |
trainer.train(resume_from_checkpoint=resume_from_checkpoint)
|
119 |
+
post_train_hooks(cfg, trainer)
|
120 |
|
121 |
LOG.info(f"Training Completed!!! Saving pre-trained model to {cfg.output_dir}")
|
122 |
|
|
|
166 |
trainer.create_model_card(model_name=cfg.output_dir.lstrip("./"))
|
167 |
|
168 |
return model, tokenizer
|
169 |
+
|
170 |
+
|
171 |
+
def pretrain_hooks(cfg, trainer):
|
172 |
+
"""
|
173 |
+
Run hooks right before kicking off the training
|
174 |
+
:param cfg:
|
175 |
+
:param trainer:
|
176 |
+
:return:
|
177 |
+
"""
|
178 |
+
neft_embeddings.pretrain_hook(cfg, trainer)
|
179 |
+
|
180 |
+
|
181 |
+
def post_train_hooks(cfg, trainer):
|
182 |
+
"""
|
183 |
+
Run hooks right after training completes
|
184 |
+
:param cfg:
|
185 |
+
:param trainer:
|
186 |
+
:return:
|
187 |
+
"""
|
188 |
+
neft_embeddings.post_train_hook(cfg, trainer)
|
src/axolotl/utils/models.py
CHANGED
@@ -180,26 +180,6 @@ def load_model(
|
|
180 |
LOG.info("patching with flash attention")
|
181 |
replace_mistral_attn_with_flash_attn(packed=cfg.sample_packing)
|
182 |
|
183 |
-
if cfg.is_llama_derived_model and cfg.noisy_embedding_alpha:
|
184 |
-
from axolotl.monkeypatch.llama_embeddings_hijack import (
|
185 |
-
replace_llama_embeddings_with_uniform_distribution,
|
186 |
-
)
|
187 |
-
|
188 |
-
LOG.info("patching with noisy embeddings")
|
189 |
-
replace_llama_embeddings_with_uniform_distribution(
|
190 |
-
noise_alpha=cfg.noisy_embedding_alpha
|
191 |
-
)
|
192 |
-
|
193 |
-
if cfg.is_mistral_derived_model and cfg.noisy_embedding_alpha:
|
194 |
-
from axolotl.monkeypatch.mistral_embeddings_hijack import (
|
195 |
-
replace_mistral_embeddings_with_uniform_distribution,
|
196 |
-
)
|
197 |
-
|
198 |
-
LOG.info("patching with noisy embeddings")
|
199 |
-
replace_mistral_embeddings_with_uniform_distribution(
|
200 |
-
noise_alpha=cfg.noisy_embedding_alpha
|
201 |
-
)
|
202 |
-
|
203 |
if cfg.is_llama_derived_model and cfg.xpos_rope:
|
204 |
from axolotl.monkeypatch.xpos_rope_llama_monkey_patch import (
|
205 |
replace_llama_rope_with_xpos_rope,
|
|
|
180 |
LOG.info("patching with flash attention")
|
181 |
replace_mistral_attn_with_flash_attn(packed=cfg.sample_packing)
|
182 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
183 |
if cfg.is_llama_derived_model and cfg.xpos_rope:
|
184 |
from axolotl.monkeypatch.xpos_rope_llama_monkey_patch import (
|
185 |
replace_llama_rope_with_xpos_rope,
|