| library_name: peft | |
| base_model: mistralai/Mistral-7B-v0.1 | |
| pipeline_tag: text-generation | |
| Description: Grammar and syntax acceptability\ | |
| Original dataset: https://huggingface.co/datasets/glue/viewer/cola \ | |
| ---\ | |
| Try querying this adapter for free in Lora Land at https://predibase.com/lora-land! \ | |
| The adapter_category is Academic Benchmarks and the name is Linguistic Acceptability (CoLA)\ | |
| ---\ | |
| Sample input: Determine if the sentence below is syntactically and semantically correct. If it is syntactically and semantically correct, respond "1". Otherwise, respond "0".\n\nSentence: Every senator seems to become more corrupt, as he talks to more lobbyists.\n\nLabel: \ | |
| ---\ | |
| Sample output: 1\ | |
| ---\ | |
| Try using this adapter yourself! | |
| ``` | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| model_id = "mistralai/Mistral-7B-v0.1" | |
| peft_model_id = "predibase/glue_cola" | |
| model = AutoModelForCausalLM.from_pretrained(model_id) | |
| model.load_adapter(peft_model_id) | |
| ``` |