JuIm
/

ProGemma

Text Generation

Generated from Trainer

text-generation-inference

Model card Files Files and versions

Metrics Training metrics Community

JuIm commited on Jul 28, 2024

Commit

1b939b2

·

verified ·

1 Parent(s): a653333

Update README.md

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -31,12 +31,12 @@ tokenizer = AutoTokenizer.from_pretrained("JuIm/Amino-Acid-Sequence-Tokenizer")
 progemma = pipeline("text-generation", model=model, tokenizer=tokenizer)
-sequence = progemma("<bos>", max_length=150, do_sample=True, top_k=950, repetition_penalty=1.2, num_return_sequences=1, eos_token_id=21)
 for i in sequence:
   print(sequence)
-'<bos>MLSLFSWFENKLDKTLKKISRIELFRKKITEVICDEHIYVMKPPFSEKTTLTREGYECGSRTMPNLARPDTYLLSRFKENCYGLHYTILGCSKNLLAPFGATFTSMLSVMVIFIFLFTKVEDFIKRCEGAGWVITEFGSTSGVPAVGPG'
 ### Framework versions

 progemma = pipeline("text-generation", model=model, tokenizer=tokenizer)
+sequence = progemma('<bos', max_length=150, do_sample=True, top_k=950, repetition_penalty=1.2, num_return_sequences=1, eos_token_id=21)
 for i in sequence:
   print(sequence)
+'MLSLFSWFENKLDKTLKKISRIELFRKKITEVICDEHIYVMKPPFSEKTTLTREGYECGSRTMPNLARPDTYLLSRFKENCYGLHYTILGCSKNLLAPFGATFTSMLSVMVIFIFLFTKVEDFIKRCEGAGWVITEFGSTSGVPAVGPG'
 ### Framework versions