Commit
•
efb8eb5
1
Parent(s):
c62e9df
Update README.md
Browse files
README.md
CHANGED
|
@@ -26,7 +26,7 @@ nf4_config = BitsAndBytesConfig(
|
|
| 26 |
bnb_4bit_compute_dtype=getattr(torch, TORCH_DTYPE)
|
| 27 |
)
|
| 28 |
|
| 29 |
-
tokenizer = AutoTokenizer.from_pretrained('mesolitica/malaysian-mistral-3B-4096')
|
| 30 |
model = AutoModelForCausalLM.from_pretrained(
|
| 31 |
'mesolitica/malaysian-mistral-3B-4096',
|
| 32 |
use_flash_attention_2 = True,
|
|
@@ -34,7 +34,6 @@ model = AutoModelForCausalLM.from_pretrained(
|
|
| 34 |
)
|
| 35 |
prompt = '<s>nama saya'
|
| 36 |
inputs = tokenizer([prompt], return_tensors='pt', add_special_tokens=False).to('cuda')
|
| 37 |
-
inputs.pop('token_type_ids')
|
| 38 |
|
| 39 |
generate_kwargs = dict(
|
| 40 |
inputs,
|
|
|
|
| 26 |
bnb_4bit_compute_dtype=getattr(torch, TORCH_DTYPE)
|
| 27 |
)
|
| 28 |
|
| 29 |
+
tokenizer = AutoTokenizer.from_pretrained('mesolitica/malaysian-mistral-3B-4096', model_input_names = ['input_ids'])
|
| 30 |
model = AutoModelForCausalLM.from_pretrained(
|
| 31 |
'mesolitica/malaysian-mistral-3B-4096',
|
| 32 |
use_flash_attention_2 = True,
|
|
|
|
| 34 |
)
|
| 35 |
prompt = '<s>nama saya'
|
| 36 |
inputs = tokenizer([prompt], return_tensors='pt', add_special_tokens=False).to('cuda')
|
|
|
|
| 37 |
|
| 38 |
generate_kwargs = dict(
|
| 39 |
inputs,
|