Update README.md
Browse files
README.md
CHANGED
@@ -38,14 +38,12 @@ import torch
|
|
38 |
model_name = "mistralai/Mistral-7B-Instruct-v0.3"
|
39 |
peft_model_id = "date3k2/mistral-v0.3-alpaca-vi-v1"
|
40 |
|
41 |
-
bnb_config = BitsAndBytesConfig(
|
42 |
-
load_in_8bit= True
|
43 |
-
)
|
44 |
model = AutoModelForCausalLM.from_pretrained(
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
)
|
50 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
51 |
|
@@ -54,9 +52,12 @@ model.load_adapter(peft_model_id)
|
|
54 |
device = "cuda"
|
55 |
|
56 |
messages = [
|
57 |
-
{
|
58 |
-
|
59 |
-
|
|
|
|
|
|
|
60 |
]
|
61 |
|
62 |
encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
|
@@ -66,6 +67,7 @@ model_inputs = encodeds.to(device)
|
|
66 |
generated_ids = model.generate(model_inputs, max_new_tokens=500, do_sample=True)
|
67 |
decoded = tokenizer.batch_decode(generated_ids)
|
68 |
print(decoded[0])
|
|
|
69 |
```
|
70 |
|
71 |
The following hyperparameters were used during training:
|
|
|
38 |
model_name = "mistralai/Mistral-7B-Instruct-v0.3"
|
39 |
peft_model_id = "date3k2/mistral-v0.3-alpaca-vi-v1"
|
40 |
|
41 |
+
bnb_config = BitsAndBytesConfig(load_in_8bit=True)
|
|
|
|
|
42 |
model = AutoModelForCausalLM.from_pretrained(
|
43 |
+
model_name,
|
44 |
+
quantization_config=bnb_config,
|
45 |
+
device_map="auto",
|
46 |
+
trust_remote_code=True,
|
47 |
)
|
48 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
49 |
|
|
|
52 |
device = "cuda"
|
53 |
|
54 |
messages = [
|
55 |
+
{
|
56 |
+
"role": "user",
|
57 |
+
"content": """You are a helpful Vietnamese AI chatbot. Below is an instruction that describes a task. Write a response that appropriately completes the request. Your response should be in Vietnamese.
|
58 |
+
Instruction:
|
59 |
+
Viết công thức để nấu một món ngon từ thịt bò.""",
|
60 |
+
},
|
61 |
]
|
62 |
|
63 |
encodeds = tokenizer.apply_chat_template(messages, return_tensors="pt")
|
|
|
67 |
generated_ids = model.generate(model_inputs, max_new_tokens=500, do_sample=True)
|
68 |
decoded = tokenizer.batch_decode(generated_ids)
|
69 |
print(decoded[0])
|
70 |
+
|
71 |
```
|
72 |
|
73 |
The following hyperparameters were used during training:
|