Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -32,8 +32,7 @@ model_path = "ov"
|
|
32 |
'''
|
33 |
hf_hub.snapshot_download(model_id, local_dir=model_path)
|
34 |
'''
|
35 |
-
|
36 |
-
pipe = ov_genai.LLMPipeline(model_path, "CPU")
|
37 |
|
38 |
config = ov_genai.GenerationConfig()
|
39 |
config.max_new_tokens = 4096
|
@@ -41,7 +40,7 @@ config.top_p = 0.9;
|
|
41 |
config.top_k = 30;
|
42 |
|
43 |
|
44 |
-
|
45 |
model_id,
|
46 |
export=True,
|
47 |
quantization_config=OVWeightQuantizationConfig(bits=4),
|
@@ -49,7 +48,10 @@ pipe = OVModelForCausalLM.from_pretrained(
|
|
49 |
|
50 |
|
51 |
# save the model after optimization
|
52 |
-
|
|
|
|
|
|
|
53 |
|
54 |
#pipe.start_chat()
|
55 |
|
|
|
32 |
'''
|
33 |
hf_hub.snapshot_download(model_id, local_dir=model_path)
|
34 |
'''
|
35 |
+
|
|
|
36 |
|
37 |
config = ov_genai.GenerationConfig()
|
38 |
config.max_new_tokens = 4096
|
|
|
40 |
config.top_k = 30;
|
41 |
|
42 |
|
43 |
+
model = OVModelForCausalLM.from_pretrained(
|
44 |
model_id,
|
45 |
export=True,
|
46 |
quantization_config=OVWeightQuantizationConfig(bits=4),
|
|
|
48 |
|
49 |
|
50 |
# save the model after optimization
|
51 |
+
model.save_pretrained(model_path)
|
52 |
+
|
53 |
+
|
54 |
+
pipe = ov_genai.LLMPipeline(model_path, "CPU")
|
55 |
|
56 |
#pipe.start_chat()
|
57 |
|