f

app.py

@@ -1,27 +1,85 @@
 import gradio as gr
-from transformers import AutoTokenizer, AutoModel
+import torch
+from transformers import AutoTokenizer, AutoModelForCausalLM
 
-# Load the model and tokenizer with `trust_remote_code=True`
-model_name = "deepseek-ai/DeepSeek-R1"
-tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
-model = AutoModel.from_pretrained(model_name, trust_remote_code=True)
+# ---------------------------------------------------------------------------
+# 1) Load the model and tokenizer
+# ---------------------------------------------------------------------------
+# If you want to load in 8-bit or 4-bit precision with bitsandbytes, 
+# uncomment and install bitsandbytes, and set load_in_8bit=True or load_in_4bit=True.
+# For example:
+#
+# from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
+# bnb_config = BitsAndBytesConfig(
+#     load_in_4bit=True,    # or load_in_8bit=True
+#     bnb_4bit_compute_dtype=torch.float16,  # recommended compute dtype
+#     bnb_4bit_use_double_quant=True,        # optional
+#     bnb_4bit_quant_type='nf4',             # optional
+# )
+#
+# model = AutoModelForCausalLM.from_pretrained(
+#     "cheberle/autotrain-35swc-b4r9z",
+#     quantization_config=bnb_config,
+#     device_map="auto",
+#     trust_remote_code=True
+# )
+# tokenizer = AutoTokenizer.from_pretrained("cheberle/autotrain-35swc-b4r9z", trust_remote_code=True)
 
-def classify_text(input_text):
-    # Tokenize input
-    inputs = tokenizer(input_text, return_tensors="pt")
-    # Get model output
-    outputs = model(**inputs)
-    probabilities = outputs.logits.softmax(dim=-1).detach().numpy()
-    return {f"Class {i}": prob for i, prob in enumerate(probabilities[0])}
-
-# Create Gradio interface
-interface = gr.Interface(
-    fn=classify_text,
-    inputs=gr.Textbox(label="Enter Text"),
-    outputs=gr.Label(label="Class Probabilities"),
-    title="DeepSeek-R1 Text Classification",
-    description="A text classification app powered by DeepSeek-R1."
+# For a standard FP16 or FP32 load (no bitsandbytes):
+model = AutoModelForCausalLM.from_pretrained(
+    "cheberle/autotrain-35swc-b4r9z",
+    device_map="auto",           # Device automatically mapped across GPUs or CPU
+    torch_dtype=torch.float16,    # Or "auto", or float32
+    trust_remote_code=True
+)
+tokenizer = AutoTokenizer.from_pretrained(
+    "cheberle/autotrain-35swc-b4r9z",
+    trust_remote_code=True
 )
 
-# Launch the app
-interface.launch()
+# ---------------------------------------------------------------------------
+# 2) Define a text generation function
+# ---------------------------------------------------------------------------
+def generate_text(prompt):
+    # Tokenize input
+    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
+
+    # Generate output (configure generation args as needed)
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_new_tokens=128,
+            temperature=0.7,
+            top_p=0.9,
+            do_sample=True
+        )
+
+    # Decode
+    decoded = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return decoded
+
+# ---------------------------------------------------------------------------
+# 3) Create the Gradio interface
+# ---------------------------------------------------------------------------
+with gr.Blocks() as demo:
+    gr.Markdown("<h3>Demo: cheberle/autotrain-35swc-b4r9z</h3>")
+    
+    with gr.Row():
+        with gr.Column():
+            prompt_in = gr.Textbox(
+                lines=5, 
+                label="Enter your prompt",
+                placeholder="Ask something here..."
+            )
+            submit_btn = gr.Button("Generate")
+        with gr.Column():
+            output_box = gr.Textbox(lines=15, label="Model Output")
+
+    # Define what happens on button click
+    submit_btn.click(fn=generate_text, inputs=prompt_in, outputs=output_box)
+
+# ---------------------------------------------------------------------------
+# 4) Launch!
+# ---------------------------------------------------------------------------
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -1,5 +1,4 @@
 huggingface_hub==0.25.2
-transformers
 torch
-datasets
-scipy
+transformers
+gradio