ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

CE Loss patching

Browse source
This commit is contained in:
Daniel Han 2024-11-02 19:20:48 -07:00
commit 726e7c933b
2 changed files with 12 additions and 150 deletions
Download patch file Download diff file Expand all files Collapse all files

4
unsloth/kernels/__init__.py
View file

@ -14,9 +14,7 @@
from .cross_entropy_loss import (
fast_cross_entropy_loss,
patch_llama_for_causal_lm,
unpatch_llama_for_causal_lm,
patch_transformers_losses,
patch_losses,
patch_loss_function,
)
from .rms_layernorm import (

158
unsloth/kernels/cross_entropy_loss.py
View file

@ -19,6 +19,12 @@ from .utils import calculate_settings, MAX_FUSED_SIZE, triton_tanh
from transformers.models.llama.modeling_llama import logger
from packaging.version import Version
from unsloth_zoo.loss_utils import (
causal_loss_function,
transformers_losses_patcher,
patch_loss_function,
)
@triton.heuristics({
"DO_SOFTCAPPING": lambda args: args["DO_SOFTCAPPING" ],
@ -354,7 +360,7 @@ class Fast_CrossEntropyLoss(torch.autograd.Function):
pass
pass
# @torch._disable_dynamo
def fast_cross_entropy_loss(
logits,
labels,
@ -382,152 +388,10 @@ def fast_cross_entropy_loss(
n_items = torch.count_nonzero(labels != -100)
return loss.sum() / n_items
pass
if Version(torch.__version__) < Version("2.5.0"):
if (Version(torch.__version__) < Version("2.4.0")) and \
not hasattr(fast_cross_entropy_loss, "__wrapped__"):
fast_cross_entropy_loss = torch._disable_dynamo(fast_cross_entropy_loss)
pass
from transformers.models.llama.modeling_llama import (
LlamaForCausalLM,
CausalLMOutputWithPast,
Optional,
Union,
Cache,
List,
Tuple,
)
# Transformers 4.47 need Unpack, KwargsForCausalLM
try:
from transformers.models.llama.modeling_llama import Unpack, KwargsForCausalLM
except:
pass
pass
import inspect, re
function = inspect.getsource(LlamaForCausalLM.forward)
function = function.split("\n")
i = re.match(r"[ ]{1,}", function[0]).span(0)[1]
function = [x[i:] for x in function]
function = "\n".join(function)
function = function[function.find("def forward"):]
replacement = """ loss = None
logit_softcapping = getattr(self.config, "final_logit_softcapping", 0)
logit_scaling = getattr(self.config, "logit_scale", 0)
if labels is not None:
shift_logits = logits
if not hasattr(self, "extra_ignored_labels"):
# Fixes https://github.com/unslothai/unsloth/issues/10
self.extra_ignored_labels = torch.full((self.max_seq_length, 1), -100, device = "cuda:0")
pass
shift_labels = torch.hstack((labels[..., 1:], self.extra_ignored_labels[:labels.shape[0]]))
loss = fast_cross_entropy_loss(
logits = shift_logits,
labels = shift_labels,
logit_softcapping = logit_softcapping,
logit_scaling = logit_scaling,
n_items = kwargs.get("num_items_in_batch", None) or kwargs.get("n_items", None),
)
else:
if logit_scaling != 0:
if logits.requires_grad:
logits = logit_scaling * logits
else:
logits *= logit_scaling
pass
pass
if logit_softcapping != 0:
if logits.requires_grad:
logits = (1.0 / logit_softcapping) * logits
logits = torch.tanh(logits)
logits = logit_softcapping * logits
else:
logits *= (1.0 / logit_softcapping)
torch.tanh(logits, out = logits)
logits *= logit_softcapping
pass
pass
pass
"""
function = \
function[:function.find(" loss = None")] + \
replacement + \
function[ function.find(" if not return_dict"):]
function = function.replace("logits = logits.float()", "\n")
# Missed spaces
function = function.split("\n")
# Not the first one though!
function = [function[0]] + [" "*4 + x for x in function[1:]]
function = "\n".join(function)
function = f"class Unsloth_LlamaForCausalLM(LlamaForCausalLM):\n"\
f" {function}\n"
exec(function, globals())
del function, replacement, inspect, re
def patch_llama_for_causal_lm():
import transformers.models.llama.modeling_llama
transformers.models.llama.modeling_llama.LlamaForCausalLM = Unsloth_LlamaForCausalLM
return
pass
def unpatch_llama_for_causal_lm():
import transformers.models.llama.modeling_llama
transformers.models.llama.modeling_llama.LlamaForCausalLM = LlamaForCausalLM
return
pass
def UnslothForCausalLMLoss(
logits, labels, vocab_size: int, num_items_in_batch: int = None, ignore_index: int = -100, **kwargs
):
shift_logits = logits
shift_labels = torch.empty_like(labels)
shift_labels[..., :-1] = labels[..., 1:]
shift_labels[..., -1] = -100
loss = fast_cross_entropy_loss(
logits = shift_logits,
labels = shift_labels,
n_items = num_items_in_batch,
)
return loss
pass
if Version(torch.__version__) < Version("2.5.0"):
UnslothForCausalLMLoss = torch._disable_dynamo(UnslothForCausalLMLoss)
pass
def patch_transformers_losses():
import re
try:
import transformers.loss.loss_utils
except:
logger.warning_once("Unsloth: Cannot patch loss functions - update transformers for faster modules!")
return
pass
import transformers.modeling_utils
LOSS_MAPPING = transformers.loss.loss_utils.LOSS_MAPPING
LOSS_MAPPING["ForCausalLM"] = UnslothForCausalLMLoss
# Remove @property and @lru_cache
if hasattr(transformers.modeling_utils.PreTrainedModel.loss_function, "fget"):
transformers.modeling_utils.PreTrainedModel.loss_function = \
transformers.modeling_utils.PreTrainedModel.loss_function.fget.__wrapped__
pass
pass
def patch_loss_function(model):
try:
# model.loss_function starts as a dict to a loss fx
# We invoke it to save it
model.loss_function = model.loss_function()
except:
# Failed means we already invoked it, and we need args to the loss fx
pass
pass
return model
pass
# Patch CE Losses in transformers
patch_losses = transformers_losses_patcher(causal_loss_function(fast_cross_entropy_loss))