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

Update vision.py

Browse source
This commit is contained in:
Daniel Han 2025-03-25 23:04:11 -07:00
commit 8e6d90bec7
1 changed files with 19 additions and 19 deletions
Download patch file Download diff file Expand all files Collapse all files

38
unsloth/models/vision.py
View file

@ -373,24 +373,24 @@ class FastBaseModel:
auto_processor = AutoProcessor if auto_model is AutoModelForVision2Seq else AutoTokenizer
tokenizer = auto_processor.from_pretrained(
tokenizer_name,
padding_side = "right",
# padding_side = "right",
token = token,
)
if hasattr(tokenizer, "tokenizer"):
__tokenizer = tokenizer.tokenizer
# Add padding side as well
__tokenizer.padding_side = "right"
# Check bos, eos, pad tokens
if hasattr(__tokenizer, "bos_token"):
tokenizer.bos_token = __tokenizer.bos_token
tokenizer.bos_token_id = __tokenizer.bos_token_id
if hasattr(__tokenizer, "eos_token"):
tokenizer.eos_token = __tokenizer.eos_token
tokenizer.eos_token_id = __tokenizer.eos_token_id
if hasattr(__tokenizer, "pad_token"):
tokenizer.pad_token = __tokenizer.pad_token
tokenizer.pad_token_id = __tokenizer.pad_token_id
pass
# if hasattr(tokenizer, "tokenizer"):
# __tokenizer = tokenizer.tokenizer
# # Add padding side as well
# __tokenizer.padding_side = "right"
# # Check bos, eos, pad tokens
# if hasattr(__tokenizer, "bos_token"):
# tokenizer.bos_token = __tokenizer.bos_token
# tokenizer.bos_token_id = __tokenizer.bos_token_id
# if hasattr(__tokenizer, "eos_token"):
# tokenizer.eos_token = __tokenizer.eos_token
# tokenizer.eos_token_id = __tokenizer.eos_token_id
# if hasattr(__tokenizer, "pad_token"):
# tokenizer.pad_token = __tokenizer.pad_token
# tokenizer.pad_token_id = __tokenizer.pad_token_id
# pass
# Fix other stuff like BnB compute data types
# model, tokenizer = patch_model_and_tokenizer(
# model,
@ -414,9 +414,9 @@ class FastBaseModel:
patch_gradient_accumulation_fix(Trainer)
# Save tokenizer for inference purposes
tokenizer.padding_side = "left" # Force inference
if hasattr(tokenizer, "tokenizer"):
tokenizer.tokenizer.padding_side = "left" # Force inference
# tokenizer.padding_side = "left" # Force inference
# if hasattr(tokenizer, "tokenizer"):
# tokenizer.tokenizer.padding_side = "left" # Force inference
m = model
while hasattr(m, "model"):
m.max_seq_length = max_seq_length