Spaces:
Running
Running
update llama model
Browse files- app.py +0 -1
- llama2_response_mail_generator.py +4 -0
app.py
CHANGED
|
@@ -78,7 +78,6 @@ print('finished download...')
|
|
| 78 |
# Initialize the Llama model with appropriate settings for GPU
|
| 79 |
lcpp_llm = Llama(
|
| 80 |
model_path=model_path_llama,
|
| 81 |
-
n_threads=2, # CPU cores to use
|
| 82 |
n_batch=512, # Batch size for processing; adjust as per your VRAM capacity
|
| 83 |
)
|
| 84 |
|
|
|
|
| 78 |
# Initialize the Llama model with appropriate settings for GPU
|
| 79 |
lcpp_llm = Llama(
|
| 80 |
model_path=model_path_llama,
|
|
|
|
| 81 |
n_batch=512, # Batch size for processing; adjust as per your VRAM capacity
|
| 82 |
)
|
| 83 |
|
llama2_response_mail_generator.py
CHANGED
|
@@ -2,7 +2,11 @@ from huggingface_hub import hf_hub_download
|
|
| 2 |
|
| 3 |
from llama_cpp import Llama
|
| 4 |
|
|
|
|
|
|
|
| 5 |
|
|
|
|
|
|
|
| 6 |
|
| 7 |
# Initialize the Llama model with appropriate settings for GPU
|
| 8 |
lcpp_llm = Llama(
|
|
|
|
| 2 |
|
| 3 |
from llama_cpp import Llama
|
| 4 |
|
| 5 |
+
model_name_or_path = "TheBloke/Llama-2-13B-chat-GGML"
|
| 6 |
+
model_basename = "llama-2-13b-chat.ggmlv3.q5_1.bin" # The model is in bin format
|
| 7 |
|
| 8 |
+
# Download the model file
|
| 9 |
+
model_path = hf_hub_download(repo_id=model_name_or_path, filename=model_basename)
|
| 10 |
|
| 11 |
# Initialize the Llama model with appropriate settings for GPU
|
| 12 |
lcpp_llm = Llama(
|