Upload 4 files

client.py

@@ -0,0 +1,16 @@
+import requests
+
+# Send prompt to the tokenizer service
+prompt = "Once upon a time"
+tokenizer_response = requests.post('http://localhost:5001/tokenize', json={'prompt': prompt})
+input_ids = tokenizer_response.json()['input_ids']
+
+# Send tokenized input to the model service
+model_response = requests.post('http://localhost:5002/generate', json={'input_ids': input_ids})
+output_ids = model_response.json()['output_ids']
+
+# Send output IDs to the decoder service
+decoder_response = requests.post('http://localhost:5003/decode', json={'output_ids': output_ids})
+generated_text = decoder_response.json()['generated_text']
+
+print("Generated Text:", generated_text)

decode.py

@@ -0,0 +1,20 @@
+from flask import Flask, request, jsonify
+from transformers import AutoTokenizer
+
+app = Flask(__name__)
+
+# Load the tokenizer for decoding
+tokenizer = AutoTokenizer.from_pretrained("verbalyze/Text2Text_Conversation_Pretrained_V2__model")
+
+@app.route('/decode', methods=['POST'])
+def decode():
+    data = request.json
+    output_ids = data.get('output_ids')
+
+    # Decode the output IDs
+    generated_text = tokenizer.decode(output_ids[0], skip_special_tokens=True)
+
+    return jsonify({'generated_text': generated_text})
+
+if __name__ == '__main__':
+    app.run(host='0.0.0.0', port=5003)

model.py

@@ -0,0 +1,21 @@
+from flask import Flask, request, jsonify
+from transformers import AutoModelForCausalLM
+import torch
+
+app = Flask(__name__)
+
+# Load the model
+model = AutoModelForCausalLM.from_pretrained("verbalyze/Text2Text_Conversation_Pretrained_V2__model")
+
+@app.route('/generate', methods=['POST'])
+def generate():
+    data = request.json
+    input_ids = torch.tensor(data.get('input_ids'))
+
+    # Generate text from input IDs
+    output_ids = model.generate(input_ids, max_length=50)
+
+    return jsonify({'output_ids': output_ids.tolist()})
+
+if __name__ == '__main__':
+    app.run(host='0.0.0.0', port=5002)

tokenizer.py

@@ -0,0 +1,23 @@
+from flask import Flask, request, jsonify
+from transformers import AutoTokenizer
+
+app = Flask(__name__)
+
+# Load tokenizer
+tokenizer = AutoTokenizer.from_pretrained("verbalyze/Text2Text_Conversation_Pretrained_V2__model")
+
+@app.route('/tokenize', methods=['POST'])
+def tokenize():
+    data = request.json
+    prompt = data.get("prompt", "")
+
+    if not prompt:
+        return jsonify({"error": "No prompt provided"}), 400
+
+    # Tokenize the prompt
+    input_ids = tokenizer(prompt, return_tensors='pt').input_ids.tolist()
+
+    return jsonify({"input_ids": input_ids})
+
+if __name__ == '__main__':
+    app.run(host='0.0.0.0', port=5001)