Upload folder using huggingface_hub

README.md

@@ -1,12 +1,12 @@
+
 ---
-title: Llm Hf Transformers Main
-emoji: 📈
-colorFrom: red
-colorTo: green
+title: llm_hf_transformers_main 
+emoji: 🔥
+colorFrom: indigo
+colorTo: indigo
 sdk: gradio
 sdk_version: 5.7.0
-app_file: app.py
+app_file: run.py
 pinned: false
+hf_oauth: true
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

requirements.txt

@@ -0,0 +1,4 @@
+gradio-client @ git+https://github.com/gradio-app/gradio@afd75def9e979d9b255a6d22d33a3aee63b6f225#subdirectory=client/python
+https://gradio-pypi-previews.s3.amazonaws.com/afd75def9e979d9b255a6d22d33a3aee63b6f225/gradio-5.7.0-py3-none-any.whl
+transformers>=4.46.0
+torch>=2.3.1

run.ipynb

@@ -0,0 +1 @@
+{"cells": [{"cell_type": "markdown", "id": "302934307671667531413257853548643485645", "metadata": {}, "source": ["# Gradio Demo: llm_hf_transformers"]}, {"cell_type": "code", "execution_count": null, "id": "272996653310673477252411125948039410165", "metadata": {}, "outputs": [], "source": ["!pip install -q gradio transformers>=4.46.0 torch>=2.3.1 "]}, {"cell_type": "code", "execution_count": null, "id": "288918539441861185822528903084949547379", "metadata": {}, "outputs": [], "source": ["from transformers import AutoModelForCausalLM, AutoTokenizer\n", "import gradio as gr\n", "\n", "checkpoint = \"HuggingFaceTB/SmolLM2-135M-Instruct\"\n", "device = \"cpu\"  # \"cuda\" or \"cpu\"\n", "tokenizer = AutoTokenizer.from_pretrained(checkpoint)\n", "model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)\n", "\n", "def predict(message, history):\n", "    history.append({\"role\": \"user\", \"content\": message})\n", "    input_text = tokenizer.apply_chat_template(history, tokenize=False)\n", "    inputs = tokenizer.encode(input_text, return_tensors=\"pt\").to(device)  # type: ignore\n", "    outputs = model.generate(inputs, max_new_tokens=100, temperature=0.2, top_p=0.9, do_sample=True)\n", "    decoded = tokenizer.decode(outputs[0])\n", "    response = decoded.split(\"<|im_start|>assistant\\n\")[-1].split(\"<|im_end|>\")[0]\n", "    return response\n", "\n", "demo = gr.ChatInterface(predict, type=\"messages\")\n", "\n", "if __name__ == \"__main__\":\n", "    demo.launch()\n"]}], "metadata": {}, "nbformat": 4, "nbformat_minor": 5}

run.py

@@ -0,0 +1,21 @@
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import gradio as gr
+
+checkpoint = "HuggingFaceTB/SmolLM2-135M-Instruct"
+device = "cpu"  # "cuda" or "cpu"
+tokenizer = AutoTokenizer.from_pretrained(checkpoint)
+model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)
+
+def predict(message, history):
+    history.append({"role": "user", "content": message})
+    input_text = tokenizer.apply_chat_template(history, tokenize=False)
+    inputs = tokenizer.encode(input_text, return_tensors="pt").to(device)  # type: ignore
+    outputs = model.generate(inputs, max_new_tokens=100, temperature=0.2, top_p=0.9, do_sample=True)
+    decoded = tokenizer.decode(outputs[0])
+    response = decoded.split("<|im_start|>assistant\n")[-1].split("<|im_end|>")[0]
+    return response
+
+demo = gr.ChatInterface(predict, type="messages")
+
+if __name__ == "__main__":
+    demo.launch()