Update README.md
Browse files
README.md
CHANGED
@@ -7,7 +7,7 @@ pipeline_tag: text2text-generation
|
|
7 |
tags:
|
8 |
- text-generation-inference
|
9 |
---
|
10 |
-
#
|
11 |
|
12 |
<!-- Provide a quick summary of what the model is/does. -->
|
13 |
```FrancescoPeriti/Llama3Dictionary``` is a fine-tuned version of the ```meta-llama/Meta-Llama-3-8B-Instruct```.
|
@@ -52,8 +52,8 @@ ft_model_name = "FrancescoPeriti/Llama3Dictionary" # fine-tuned model
|
|
52 |
|
53 |
# load models
|
54 |
chat_model = AutoModelForCausalLM.from_pretrained(model_name, device_map='auto')
|
55 |
-
|
56 |
-
|
57 |
|
58 |
# load tokenizer
|
59 |
tokenizer = AutoTokenizer.from_pretrained(
|
@@ -127,7 +127,7 @@ with torch.no_grad():
|
|
127 |
for k in ['input_ids', 'attention_mask']:
|
128 |
model_input[k] = torch.tensor(batch[k]).to('cuda')
|
129 |
|
130 |
-
output_ids =
|
131 |
max_length = max_length * batch_size,
|
132 |
forced_eos_token_id = eos_tokens,
|
133 |
max_time = max_time * batch_size,
|
|
|
7 |
tags:
|
8 |
- text-generation-inference
|
9 |
---
|
10 |
+
# Llama3Dictionary
|
11 |
|
12 |
<!-- Provide a quick summary of what the model is/does. -->
|
13 |
```FrancescoPeriti/Llama3Dictionary``` is a fine-tuned version of the ```meta-llama/Meta-Llama-3-8B-Instruct```.
|
|
|
52 |
|
53 |
# load models
|
54 |
chat_model = AutoModelForCausalLM.from_pretrained(model_name, device_map='auto')
|
55 |
+
lama3dictionary = PeftModel.from_pretrained(chat_model, ft_model_name)
|
56 |
+
lama3dictionary.eval()
|
57 |
|
58 |
# load tokenizer
|
59 |
tokenizer = AutoTokenizer.from_pretrained(
|
|
|
127 |
for k in ['input_ids', 'attention_mask']:
|
128 |
model_input[k] = torch.tensor(batch[k]).to('cuda')
|
129 |
|
130 |
+
output_ids = lama3dictionary.generate(**model_input,
|
131 |
max_length = max_length * batch_size,
|
132 |
forced_eos_token_id = eos_tokens,
|
133 |
max_time = max_time * batch_size,
|