Update app.py

app.py

@@ -1,64 +1,101 @@
-import gradio as gr
 import argparse
+from tabulate import tabulate
 
 def main():
-    parser = argparse.ArgumentParser(description='Estimare capacità e latenza di un modello LLM')
-    parser.add_argument('--gpu', type=str, default='A100 80GB', help='Tipo di GPU')
-    parser.add_argument('--model', type=str, default='Llama-3-70B', help='Nome del modello')
-    parser.add_argument('--prompt_size', type=int, default=4096, help='Dimensione della promessa in token')
-    parser.add_argument('--response_size', type=int, default=256, help='Dimensione della risposta in token')
-    parser.add_argument('--concurrent_requests', type=int, default=10, help='Numero di richieste concorrenti')
+    parser = argparse.ArgumentParser(description='Your script description')
+    parser.add_argument('-g', '--num_gpu', type=int, default=1, help='Number of GPUs')
+    parser.add_argument('-p', '--prompt_sz', type=int, default=4096, help='Prompt size in tokens')
+    parser.add_argument('-r', '--response_sz', type=int, default=256, help='Response size in tokens')
+    parser.add_argument('-c', '--n_concurrent_req', type=int, default=10, help='Number of concurrent requests')
+    parser.add_argument('-w', '-cw', '--ctx_window', type=int, default=1024, help='Average context window')
 
     args = parser.parse_args()
 
-    gpu_specs = {
-        'A100 80GB': {'tflops': 312, 'memory_gb': 80, 'bandwidth': 1935},
-        'H100 SXM': {'tflops': 1979, 'memory_gb': 80, 'bandwidth': 3350},
-    }
-
-    model_specs = {
-        'Llama-3-70B': {'params_billion': 70, 'd_model': 8192, 'n_layers': 80},
-        'Llama-3-8B': {'params_billion': 8, 'd_model': 4096, 'n_layers': 32},
-    }
-
-    def estimate_llm_capacity(model_name, gpu_name, prompt_size, response_size, concurrent_requests):
-        gpu = gpu_specs[gpu_name]
-        model = model_specs[model_name]
-
-        kv_cache_tokens = (gpu['tflops'] * concurrent_requests) // (model['params_billion'] * 2)
-        prefill_time_ms = (model['params_billion'] * 2) / (gpu['tflops'] * concurrent_requests) * 1000
-        generation_time_ms = (model['params_billion'] * 2) / (gpu['bandwidth'] * concurrent_requests) * 1000
-        estimated_response_time = (prompt_size * prefill_time_ms + response_size * generation_time_ms) / 1000
-
-        return f"""
-Modello: {model_name}
-GPU: {gpu_name}
-KV Cache Tokens: {kv_cache_tokens:.0f}
-Prefill Time: {prefill_time_ms:.2f} ms
-Generation Time: {generation_time_ms:.2f} ms
-Estimated Response Time: {estimated_response_time:.2f} s
-"""
-
-    with gr.Blocks() as demo:
-        gr.Markdown("# Estimare Capacità e Latenza LLM")
-        
-        gpu_dropdown = gr.Dropdown(choices=['A100 80GB', 'H100 SXM'], label="Tipo di GPU", value='A100 80GB')
-        model_dropdown = gr.Dropdown(choices=['Llama-3-70B', 'Llama-3-8B'], label="Nome del Modello", value='Llama-3-70B')
-        prompt_size = gr.Slider(minimum=1, maximum=8192, label="Dimensione della Promessa", value=4096)
-        response_size = gr.Slider(minimum=1, maximum=8192, label="Dimensione della Risposta", value=256)
-        concurrent_requests = gr.Slider(minimum=1, maximum=100, label="Richieste Concorrenti", value=10)
-
-        table = gr.Textbox()
-
-        submit_button = gr.Button("Estimare")
-
-        submit_button.click(
-            fn=lambda gpu=gpu_dropdown.value, model=model_dropdown.value, 
-            prompt_size=prompt_size.value, response_size=response_size.value, 
-            concurrent_requests=concurrent_requests.value:
-                estimate_llm_capacity(model, gpu, prompt_size, response_size, concurrent_requests),
-            inputs=[gpu_dropdown, model_dropdown, prompt_size, response_size, concurrent_requests],
-            outputs=[table]
-        )
-
-    demo.launch()
+    num_gpu = args.num_gpu
+    prompt_size = args.prompt_sz
+    response_size = args.response_sz
+    n_concurrent_request = args.n_concurrent_req
+    avg_context_window = args.ctx_window
+
+    # Print input
+    print(f" num_gpu = {num_gpu}, prompt_size = {prompt_size} tokens, response_size = {response_size} tokens")
+    print(f" n_concurrent_request = {n_concurrent_request}, avg_context_window = {avg_context_window} tokens")
+
+    # Define variables
+    gpu_specs = [
+        {"name": "A10", "fp16_tflops": 125, "memory_gb": 24, "memory_bandwidth_gbps": 600},
+        {"name": "A30", "fp16_tflops": 330, "memory_gb": 24, "memory_bandwidth_gbps": 933},
+        {"name": "L40", "fp16_tflops": 181, "memory_gb": 48, "memory_bandwidth_gbps": 864},
+        {"name": "L40s", "fp16_tflops": 362, "memory_gb": 48, "memory_bandwidth_gbps": 864},
+        {"name": "A100 40 GB", "fp16_tflops": 312, "memory_gb": 40, "memory_bandwidth_gbps": 1555},
+        {"name": "A100 40 GB SXM", "fp16_tflops": 312, "memory_gb": 40, "memory_bandwidth_gbps": 1555},
+        {"name": "A100 80 GB PCIe", "fp16_tflops": 312, "memory_gb": 80, "memory_bandwidth_gbps": 1935},
+        {"name": "A100 80 GB SXM", "fp16_tflops": 312, "memory_gb": 80, "memory_bandwidth_gbps": 2039},
+        {"name": "H100 PCIe", "fp16_tflops": 1513, "memory_gb": 80, "memory_bandwidth_gbps": 2000},
+        {"name": "H100 SXM", "fp16_tflops": 1979, "memory_gb": 80, "memory_bandwidth_gbps": 3350},
+        {"name": "H100 NVL", "fp16_tflops": 3958, "memory_gb": 188, "memory_bandwidth_gbps": 7800}
+        # Add or comment out GPU types as needed
+    ]
+
+    model_specs = [
+        {"name": "Llama-3-8B", "params_billion": 8, "d_model": 4096, "n_heads": 32, "n_layers": 32, "max_context_window": 8192, "d_head": 128},
+        {"name": "Llama-3-70B", "params_billion": 70, "d_model": 8192, "n_heads": 64, "n_layers": 80, "max_context_window": 8192, "d_head": 128},
+        {"name": "Llama-3.1-8B", "params_billion": 8, "d_model": 4096, "n_heads": 32, "n_layers": 32, "max_context_window": 131072, "d_head": 128},
+        {"name": "Llama-3.1-70B", "params_billion": 70, "d_model": 8192, "n_heads": 64, "n_layers": 80, "max_context_window": 131072, "d_head": 128},
+        {"name": "Mistral-7B-v0.3", "params_billion": 7, "d_model": 4096, "n_heads": 32, "n_layers": 32, "max_context_window": 32768, "d_head": 128},
+        {"name": "Falcon-7B", "params_billion": 7, "d_model": 4544, "n_heads": 71, "n_layers": 32, "max_context_window": 2048, "d_head": 64},
+        {"name": "Falcon-40B", "params_billion": 40, "d_model": 8192, "n_heads": 128, "n_layers": 60, "max_context_window": 2048, "d_head": 64},
+        {"name": "Falcon-180B", "params_billion": 180, "d_model": 14848, "n_heads": 232, "n_layers": 80, "max_context_window": 2048, "d_head": 64}
+        # Add or comment out model specifications as needed
+    ]
+
+    BYTES_IN_GB = 1_073_741_824  # 1 GB = 1,073,741,824 bytes
+
+    def calc_kv_cache_size_per_token(n_layers, d_model):
+        return 2 * 2 * n_layers * d_model / BYTES_IN_GB  # GB/token
+
+    def calc_memory_footprint(model_spec, n_concurrent_request, avg_context_window):
+        kv_cache_size_per_token = calc_kv_cache_size_per_token(model_spec["n_layers"], model_spec["d_model"])
+        target_gpu_mem = kv_cache_size_per_token * avg_context_window * n_concurrent_request + model_spec["params_billion"] * 2
+        return target_gpu_mem
+
+    print(f"\n******************** Estimate LLM Memory Footprint ********************")
+    memory_footprint_table = []
+    for model_spec in model_specs:
+        kv_cache_size_per_token = calc_kv_cache_size_per_token(model_spec["n_layers"], model_spec["d_model"])
+        memory_footprint = calc_memory_footprint(model_spec, n_concurrent_request, avg_context_window)
+        memory_footprint_table.append([model_spec['name'], f"{kv_cache_size_per_token:.6f} GiB/token", f"{memory_footprint:.2f} GB"])
+    print(tabulate(memory_footprint_table, headers=['Model', 'KV Cache Size per Token', 'Memory Footprint'], tablefmt='orgtbl'))
+
+    def calc_kv_cache_tokens(num_gpu, gpu_memory_gb, model_params_billion, kv_cache_size):
+        result = (num_gpu * gpu_memory_gb - 2 * model_params_billion) / kv_cache_size
+        return result if result >= 0 else "OOM"
+
+    def calc_prefill_time_per_token(num_gpu, model_params_billion, fp16_tflops):
+        result = (2 * model_params_billion / num_gpu) / fp16_tflops
+        return result if result >= 0 else "OOM"
+
+    def calc_generation_time_per_token(num_gpu, model_params_billion, memory_bandwidth_gbps):
+        result = (2 * model_params_billion / num_gpu) / memory_bandwidth_gbps * 1000
+        return result if result >= 0 else "OOM"
+
+    def calc_estimated_response_time(prefill_time, generation_time, prompt_size, response_size):
+        if isinstance(prefill_time, str) or isinstance(generation_time, str):  # Check if any are "NA"
+            return "OOM"
+        return (prompt_size * prefill_time + response_size * generation_time) / 1000  # convert ms to seconds
+
+    print(f"\n******************** Estimate LLM Capacity and Latency ******************** ")
+    capacity_latency_table = []
+    for model in model_specs:
+        # print(f"Model: {model['name']} ({model['params_billion']}B parameters)")
+        kv_cache_size = calc_kv_cache_size_per_token(model['n_layers'], model['d_model'])
+        for gpu in gpu_specs:
+            kv_cache_tokens = calc_kv_cache_tokens(num_gpu, gpu['memory_gb'], model['params_billion'], kv_cache_size)
+            prefill_time_per_token = calc_prefill_time_per_token(num_gpu, model['params_billion'], gpu['fp16_tflops'])
+            generation_time_per_token = calc_generation_time_per_token(num_gpu, model['params_billion'], gpu['memory_bandwidth_gbps'])
+            estimated_response_time = calc_estimated_response_time(prefill_time_per_token, generation_time_per_token, prompt_size, response_size)
+            capacity_latency_table.append([model['name'], gpu['name'], f"{kv_cache_tokens}", f"{prefill_time_per_token:.3f} ms", f"{generation_time_per_token:.3f} ms", f"{estimated_response_time:.1f} s"])
+    print(tabulate(capacity_latency_table, headers=['Model', 'GPU', 'KV Cache Tokens', 'Prefill Time', 'Generation Time', 'Estimated Response Time'], tablefmt='orgtbl'))
+
+if __name__ == '__main__':
+    main()