casts the prepared data to int16 (doesn't help with training memory)
Browse files- scripts/finetune.py +1 -2
- src/axolotl/datasets.py +12 -5
scripts/finetune.py
CHANGED
|
@@ -14,7 +14,6 @@ import transformers
|
|
| 14 |
import yaml
|
| 15 |
from attrdict import AttrDefault
|
| 16 |
from datasets import load_dataset, IterableDataset, Dataset, load_from_disk
|
| 17 |
-
from huggingface_hub.hf_api import DatasetInfo
|
| 18 |
from torch import nn
|
| 19 |
from transformers import (
|
| 20 |
AutoModelForCausalLM,
|
|
@@ -169,7 +168,7 @@ def load_model(base_model, base_model_config, model_type, tokenizer_type, cfg, a
|
|
| 169 |
|
| 170 |
if cfg.load_4bit:
|
| 171 |
# Scales to half
|
| 172 |
-
|
| 173 |
for n, m in model.named_modules():
|
| 174 |
if 'Autograd4bitQuantLinear' in str(type(m)) or 'Linear4bitLt' in str(type(m)):
|
| 175 |
if hasattr(m, "is_v1_model") and m.is_v1_model:
|
|
|
|
| 14 |
import yaml
|
| 15 |
from attrdict import AttrDefault
|
| 16 |
from datasets import load_dataset, IterableDataset, Dataset, load_from_disk
|
|
|
|
| 17 |
from torch import nn
|
| 18 |
from transformers import (
|
| 19 |
AutoModelForCausalLM,
|
|
|
|
| 168 |
|
| 169 |
if cfg.load_4bit:
|
| 170 |
# Scales to half
|
| 171 |
+
logging.info('Fitting 4bit scales and zeros to half')
|
| 172 |
for n, m in model.named_modules():
|
| 173 |
if 'Autograd4bitQuantLinear' in str(type(m)) or 'Linear4bitLt' in str(type(m)):
|
| 174 |
if hasattr(m, "is_v1_model") and m.is_v1_model:
|
src/axolotl/datasets.py
CHANGED
|
@@ -30,7 +30,6 @@ class TokenizedPromptDataset(IterableDataset):
|
|
| 30 |
except InvalidDataException:
|
| 31 |
pass
|
| 32 |
|
| 33 |
-
|
| 34 |
# TODO this isn't the best since it can't interleave datasets
|
| 35 |
class ConstantLengthDataset(IterableDataset):
|
| 36 |
"""
|
|
@@ -40,7 +39,6 @@ class ConstantLengthDataset(IterableDataset):
|
|
| 40 |
dataset (dataset.Dataset): Dataset with text files.
|
| 41 |
seq_length (int): Length of token sequences to return.
|
| 42 |
"""
|
| 43 |
-
|
| 44 |
def __init__(
|
| 45 |
self,
|
| 46 |
tokenizer,
|
|
@@ -52,6 +50,15 @@ class ConstantLengthDataset(IterableDataset):
|
|
| 52 |
self.datasets: List[IterableDataset] = datasets
|
| 53 |
self.seq_length = seq_length
|
| 54 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
def __iter__(self):
|
| 56 |
buffer = {"input_ids": [], "attention_mask": [], "labels": []}
|
| 57 |
buffer_len = 0
|
|
@@ -105,11 +112,11 @@ class ConstantLengthDataset(IterableDataset):
|
|
| 105 |
attention_mask.append(1)
|
| 106 |
labels.append(self.concat_token_id)
|
| 107 |
|
| 108 |
-
input_ids_with_concat = torch.tensor(input_ids, dtype=
|
| 109 |
attention_mask_with_concat = torch.tensor(
|
| 110 |
-
attention_mask, dtype=
|
| 111 |
)
|
| 112 |
-
labels_with_concat = torch.tensor(labels, dtype=
|
| 113 |
|
| 114 |
buffer["input_ids"].append(input_ids_with_concat)
|
| 115 |
buffer["attention_mask"].append(attention_mask_with_concat)
|
|
|
|
| 30 |
except InvalidDataException:
|
| 31 |
pass
|
| 32 |
|
|
|
|
| 33 |
# TODO this isn't the best since it can't interleave datasets
|
| 34 |
class ConstantLengthDataset(IterableDataset):
|
| 35 |
"""
|
|
|
|
| 39 |
dataset (dataset.Dataset): Dataset with text files.
|
| 40 |
seq_length (int): Length of token sequences to return.
|
| 41 |
"""
|
|
|
|
| 42 |
def __init__(
|
| 43 |
self,
|
| 44 |
tokenizer,
|
|
|
|
| 50 |
self.datasets: List[IterableDataset] = datasets
|
| 51 |
self.seq_length = seq_length
|
| 52 |
|
| 53 |
+
vocab_size = len(tokenizer.get_vocab())
|
| 54 |
+
|
| 55 |
+
if vocab_size <= torch.iinfo(torch.int16).max:
|
| 56 |
+
self.tokens_dtype = torch.int16
|
| 57 |
+
elif vocab_size <= torch.iinfo(torch.int32).max:
|
| 58 |
+
self.tokens_dtype = torch.int32
|
| 59 |
+
else:
|
| 60 |
+
self.tokens_dtype = torch.int64
|
| 61 |
+
|
| 62 |
def __iter__(self):
|
| 63 |
buffer = {"input_ids": [], "attention_mask": [], "labels": []}
|
| 64 |
buffer_len = 0
|
|
|
|
| 112 |
attention_mask.append(1)
|
| 113 |
labels.append(self.concat_token_id)
|
| 114 |
|
| 115 |
+
input_ids_with_concat = torch.tensor(input_ids, dtype=self.tokens_dtype)
|
| 116 |
attention_mask_with_concat = torch.tensor(
|
| 117 |
+
attention_mask, dtype=self.tokens_dtype
|
| 118 |
)
|
| 119 |
+
labels_with_concat = torch.tensor(labels, dtype=self.tokens_dtype)
|
| 120 |
|
| 121 |
buffer["input_ids"].append(input_ids_with_concat)
|
| 122 |
buffer["attention_mask"].append(attention_mask_with_concat)
|