Update README.md
Browse files
README.md
CHANGED
|
@@ -42,8 +42,8 @@ pip install transformers==4.51.0
|
|
| 42 |
```python
|
| 43 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 44 |
|
|
|
|
| 45 |
model_name = "infly/inf-query-aligner"
|
| 46 |
-
|
| 47 |
model = AutoModelForCausalLM.from_pretrained(
|
| 48 |
model_name,
|
| 49 |
torch_dtype="auto",
|
|
@@ -51,12 +51,14 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 51 |
)
|
| 52 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
| 53 |
|
|
|
|
| 54 |
prompt = "Give me a short introduction to large language model."
|
| 55 |
messages = [
|
| 56 |
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
|
| 57 |
{"role": "user", "content": prompt}
|
| 58 |
]
|
| 59 |
|
|
|
|
| 60 |
text = tokenizer.apply_chat_template(
|
| 61 |
messages,
|
| 62 |
tokenize=False,
|
|
@@ -64,6 +66,7 @@ text = tokenizer.apply_chat_template(
|
|
| 64 |
)
|
| 65 |
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
| 66 |
|
|
|
|
| 67 |
generated_ids = model.generate(
|
| 68 |
**model_inputs,
|
| 69 |
max_new_tokens=512
|
|
@@ -73,6 +76,8 @@ generated_ids = [
|
|
| 73 |
]
|
| 74 |
|
| 75 |
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
|
|
|
|
|
|
| 76 |
```
|
| 77 |
|
| 78 |
---
|
|
|
|
| 42 |
```python
|
| 43 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 44 |
|
| 45 |
+
# Load model and tokenizer
|
| 46 |
model_name = "infly/inf-query-aligner"
|
|
|
|
| 47 |
model = AutoModelForCausalLM.from_pretrained(
|
| 48 |
model_name,
|
| 49 |
torch_dtype="auto",
|
|
|
|
| 51 |
)
|
| 52 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
| 53 |
|
| 54 |
+
# Define input query
|
| 55 |
prompt = "Give me a short introduction to large language model."
|
| 56 |
messages = [
|
| 57 |
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
|
| 58 |
{"role": "user", "content": prompt}
|
| 59 |
]
|
| 60 |
|
| 61 |
+
# Apply chat template
|
| 62 |
text = tokenizer.apply_chat_template(
|
| 63 |
messages,
|
| 64 |
tokenize=False,
|
|
|
|
| 66 |
)
|
| 67 |
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
| 68 |
|
| 69 |
+
# Generate rewritten query
|
| 70 |
generated_ids = model.generate(
|
| 71 |
**model_inputs,
|
| 72 |
max_new_tokens=512
|
|
|
|
| 76 |
]
|
| 77 |
|
| 78 |
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
| 79 |
+
|
| 80 |
+
print(response)
|
| 81 |
```
|
| 82 |
|
| 83 |
---
|