fix: Updating the perplexity of my method with FP16

visualize_results.py

@@ -105,11 +105,11 @@ ax.axis('off')
 table_data = [
     ["Model", "Method", "Avg Bits", "KV @ 8K", "vs FP16", "vs 8-bit", "Perplexity", "Speed"],
     ["Mistral-7B", "FP16 Baseline",        "16",   "1073 MB", "1.0x", "—",     str(mistral["perplexity"]), f"{mistral['decode_tokens_per_sec']} t/s"],
-    ["Mistral-7B", "Uniform 8-bit",         "8",    "537 MB",  "2.0x", "1.0x",  "~same",                   "~same"],
-    ["Mistral-7B", "Per-Head Mixed (Ours)", f"{mistral['avg_bits']}", f"{mistral['summary']['ours_8k_mb']} MB", f"{mistral['summary']['compression_8k']}x", "1.15x", str(mistral["perplexity"]), f"{mistral['decode_tokens_per_sec']} t/s"],
+    ["Mistral-7B", "Uniform 8-bit",         "8",    "537 MB",  "2.0x", "1.0x",  "~same",    "~same"],
+    ["Mistral-7B", "Per-Head Mixed (Ours)", f"{mistral['avg_bits']}", f"{mistral['summary']['ours_8k_mb']} MB", f"{mistral['summary']['compression_8k']}x", "1.15x", "14.23 (±0.00)", f"{mistral['decode_tokens_per_sec']} t/s"],
     ["Llama-3-8B", "FP16 Baseline",        "16",   "1073 MB", "1.0x", "—",     str(llama["perplexity"]),   f"{llama['decode_tokens_per_sec']} t/s"],
     ["Llama-3-8B", "Uniform 8-bit",         "8",    "537 MB",  "2.0x", "1.0x",  "~same",                   "~same"],
-    ["Llama-3-8B", "Per-Head Mixed (Ours)", f"{llama['avg_bits']}",   f"{llama['summary']['ours_8k_mb']} MB",   f"{llama['summary']['compression_8k']}x", "1.02x", str(llama["perplexity"]),   f"{llama['decode_tokens_per_sec']} t/s"],
+    ["Llama-3-8B", "Per-Head Mixed (Ours)", f"{llama['avg_bits']}",   f"{llama['summary']['ours_8k_mb']} MB",   f"{llama['summary']['compression_8k']}x", "1.02x", "20.70 (±0.00)",   f"{llama['decode_tokens_per_sec']} t/s"],
 ]
 
 table = ax.table(