fixed hugging face llm

examples/example_huggingfacellm.py

@@ -1,18 +1,29 @@
 from swarms.models import HuggingfaceLLM
+import torch
 
-# Initialize with custom configuration
+try:
-custom_config = {
+    inference = HuggingfaceLLM(
-    "quantize": True,
+        model_id="gpt2",
-    "quantization_config": {"load_in_4bit": True},
+        quantize=False,
-    "verbose": True,
+        verbose=True,
-}
+    )
-inference = HuggingfaceLLM(
-    model_id="NousResearch/Nous-Hermes-2-Vision-Alpha", **custom_config
-)
 
-# Generate text based on a prompt
+    device = "cuda" if torch.cuda.is_available() else "cpu"
-prompt_text = (
+    inference.model.to(device)
-    "Create a list of known biggest risks of structural collapse with references"
+
-)
+    prompt_text = "Create a list of known biggest risks of structural collapse with references"
-generated_text = inference(prompt_text)
+    inputs = inference.tokenizer(prompt_text, return_tensors="pt").to(device)
-print(generated_text)
+    
+    generated_ids = inference.model.generate(
+        **inputs,
+        max_new_tokens=1000,  # Adjust the length of the generation
+        temperature=0.7,  # Adjust creativity
+        top_k=50,  # Limits the vocabulary considered at each step
+        pad_token_id=inference.tokenizer.eos_token_id,
+        do_sample=True  # Enable sampling to utilize temperature
+    )
+    
+    generated_text = inference.tokenizer.decode(generated_ids[0], skip_special_tokens=True)
+    print(generated_text)
+except Exception as e:
+    print(f"An error occurred: {e}")