Downloads · 30 days
8
2% of all-time downloads
WizardVicuna-pythia-1.4b-deduped
heegyu/WizardVicuna-pythia-1.4b-deduped
WizardVicuna-pythia-1.4b-deduped is a text generation model from heegyu. Use it when you need the model to write or continue text. It is set up for transformers.
Hyperparameters - 3 epoch - 1e-4 - 1e-5 with cosine lr decay - batch size 128 - max sequence length 2048 - AdamW(weigth decay=0.01, b1=0.9, b2=0.99, gradclip=1.0) - no warmup - BF16
All-time downloads
342
Public
Repo size
5.7 GB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.bin2.8 GB · 100%
At a glance
- Task
- Text Generation
- Library
- transformers
- Model type
- gpt_neox
- Access
- Public
- Created
- Aug 25, 2023
- Updated
- Aug 25, 2023
- SHA
- a729d119
Try a prompt
Datasets
Model card
From the Hugging Face model README
Hyperparameters
- 3 epoch
- 1e-4 -> 1e-5 with cosine lr decay
- batch size 128
- max sequence length 2048
- AdamW(weigth decay=0.01, b1=0.9, b2=0.99, grad_clip=1.0)
- no warmup
- BF16
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("heegyu/WizardVicuna-pythia-1.4b-deduped")
model = AutoModelForCausalLM.from_pretrained("heegyu/WizardVicuna-pythia-1.4b-deduped")
inputs = tokenizer(["Human: Hi\n\nAssistant: "], return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=16)
print(tokenizer.batch_decode(outputs, skip_special_tokens=False))
output: ['Human: Hi\n\nAssistant: Hello! How can I assist you today?<|endoftext|>']
- Task
- Text Generation
- Library
- transformers
- Type
- gpt_neox
- Created
- Aug 25, 2023
- Updated
- Aug 25, 2023