Update README.md
Browse files
README.md
CHANGED
@@ -18,28 +18,30 @@ This model was trained using AutoTrain. For more information, please visit [Auto
|
|
18 |
|
19 |
# Usage
|
20 |
|
21 |
-
```
|
|
|
|
|
22 |
|
|
|
23 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
24 |
|
25 |
-
|
|
|
26 |
|
27 |
-
|
28 |
-
|
29 |
-
model_path,
|
30 |
-
device_map="auto",
|
31 |
-
torch_dtype='auto'
|
32 |
-
).eval()
|
33 |
|
34 |
-
# Prompt content: "hi"
|
35 |
messages = [
|
36 |
-
{"role": "user", "content": "
|
37 |
]
|
|
|
|
|
|
|
38 |
|
39 |
input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
|
40 |
-
output_ids = model.generate(input_ids.to(
|
41 |
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
|
42 |
|
43 |
-
# Model response: "Hello! How can I assist you today?"
|
44 |
print(response)
|
45 |
```
|
|
|
18 |
|
19 |
# Usage
|
20 |
|
21 |
+
```bash
|
22 |
+
pip install peft
|
23 |
+
```
|
24 |
|
25 |
+
```python
|
26 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
27 |
+
from peft import AutoPeftModelForCausalLM, PeftConfig
|
28 |
|
29 |
+
model_id = "Aryan-401/phi-3-mini-4k-instruct-finetune-guanaco"
|
30 |
+
peft_model=AutoPeftModelForCausalLM.from_pretrained(model_id)
|
31 |
|
32 |
+
model = peft_model.merge_and_unload()
|
33 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
|
|
|
|
|
|
|
|
34 |
|
|
|
35 |
messages = [
|
36 |
+
{"role": "user", "content": "What is the Value of Pi?"}
|
37 |
]
|
38 |
+
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
|
39 |
+
|
40 |
+
model = model.to(device).eval()
|
41 |
|
42 |
input_ids = tokenizer.apply_chat_template(conversation=messages, tokenize=True, add_generation_prompt=True, return_tensors='pt')
|
43 |
+
output_ids = model.generate(input_ids.to(device), max_length= 1000)
|
44 |
response = tokenizer.decode(output_ids[0][input_ids.shape[1]:], skip_special_tokens=True)
|
45 |
|
|
|
46 |
print(response)
|
47 |
```
|