sahil2801 commited on
Commit
c21e0f1
1 Parent(s): 822b369

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +30 -0
README.md CHANGED
@@ -1,3 +1,33 @@
1
  ---
2
  license: bsd-3-clause
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: bsd-3-clause
3
+ metrics:
4
+ - code_eval
5
+ pipeline_tag: text-generation
6
+ tags:
7
+ - code
8
  ---
9
+ # Model Card for instruct-codegen-16B
10
+
11
+ <!-- Provide a quick summary of what the model is/does. -->
12
+
13
+ Instruct-codegen-16B is an instruction following codegen model based on [Salesforce codegen-16B-multi](https://huggingface.co/Salesforce/codegen-16B-multi) , finetuned on a dataset of 250k instruction-following samples in the alpaca format.
14
+
15
+ The data was not generated using any commercial LLM api.
16
+
17
+ The model achieves a new SoTA result of 36.1% pass@1 on the HumanEval benchmark.
18
+
19
+ ## Generation
20
+
21
+ ```python
22
+ # pip install -q transformers
23
+ from transformers import AutoModelForCausalLM, AutoTokenizer
24
+ checkpoint = "sahil2801/instruct-codegen-16B"
25
+ device = "cuda"
26
+ tokenizer = AutoTokenizer.from_pretrained(checkpoint)
27
+ model = AutoModelForCausalLM.from_pretrained(checkpoint).half().to(device)
28
+ instruction = "Write a function to scrape hacker news."
29
+ prompt = f"Below is an instruction that describes a task.\n Write a response that appropriately completes the request.\n\n ### Instruction:\n{instruction}\n\n### Response:"
30
+ inputs = tokenizer(prompt, return_tensors="pt").to(device)
31
+ outputs = model.generate(**inputs,temperature=0.3,do_sample=True)
32
+ print(tokenizer.decode(outputs[0],skip_special_tokens=True))
33
+ ```