hrezaei/nanoGPTLookAhead
Update README.md
Add format:pt to metadata
Remove unused imports
Try to fix the bug in handling device_map
Add the architectures key to config to avoid an error in loading model
Upload model.py
Upload model.py
Upload model.py
Fix arguments of model.generate for compatibility with Huggingface
Fix max_new_tokens in model.generate
Upload model.py
Upload model.py
Upload tokenizer
WandB run: uoy/owt/2spjerwi
Add model to the device given to constructor
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Add support for past key and values in attention layer
Add GPTLAConfig to be used for GPTLA model
Push model using huggingface_hub.
Update output of GPTLA for separate look ahead logits
Upload model.py
Push model using huggingface_hub.
Modify auto model registered classes manually
Delete nanogpt_model.py
Add two python files to support auto load
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Add nanogpt_model.py
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Upload tokenizer
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
Push model using huggingface_hub.
initial commit