Question Answering
Transformers
PyTorch
TensorFlow
Safetensors
PEFT
English
deberta-v2
deberta
deberta-v3
squad
squad_v2
lora
Eval Results (legacy)
Instructions to use sjrhuschlee/deberta-v3-large-squad2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sjrhuschlee/deberta-v3-large-squad2 with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "question-answering" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # pip install "transformers<5.0.0" from transformers import pipeline pipe = pipeline("question-answering", model="sjrhuschlee/deberta-v3-large-squad2")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForQuestionAnswering tokenizer = AutoTokenizer.from_pretrained("sjrhuschlee/deberta-v3-large-squad2") model = AutoModelForQuestionAnswering.from_pretrained("sjrhuschlee/deberta-v3-large-squad2", device_map="auto") - PEFT
How to use sjrhuschlee/deberta-v3-large-squad2 with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
Add TF weights
#2
by sjrhuschlee - opened
Model converted by the transformers' pt_to_tf CLI. All converted model outputs and hidden layers were validated against its PyTorch counterpart.
Maximum crossload output difference=2.289e-05; Maximum crossload hidden layer difference=6.390e-05;
Maximum conversion output difference=2.289e-05; Maximum conversion hidden layer difference=6.390e-05;
CAUTION: The maximum admissible error was manually increased to 6.5e-05!
sjrhuschlee changed pull request status to merged