1ff3d25cf63eeb7d4db79ea99406cccfb60dae0b5906cd6b07533efe2eb32261

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "/tmp/tmpcmzh3zye",
+  "_name_or_path": "/tmp/tmpu4qxllsm",
   "architectures": [
     "LlamaForCausalLM"
   ],

results.json

@@ -1,30 +1,30 @@
 {
     "base_current_gpu_type": "NVIDIA A100-PCIE-40GB",
     "base_current_gpu_total_memory": 40339.3125,
-    "base_token_generation_latency_sync": 76.9214370727539,
-    "base_token_generation_latency_async": 76.84411387890577,
-    "base_token_generation_throughput_sync": 0.01300027714061269,
-    "base_token_generation_throughput_async": 0.01301335846719298,
-    "base_token_generation_CO2_emissions": 4.107156647942229e-05,
-    "base_token_generation_energy_consumption": 0.013795784438629225,
-    "base_inference_latency_sync": 75.23112869262695,
-    "base_inference_latency_async": 74.10402297973633,
-    "base_inference_throughput_sync": 0.01329237002525532,
-    "base_inference_throughput_async": 0.013494544017852432,
-    "base_inference_CO2_emissions": 4.120221930394127e-05,
-    "base_inference_energy_consumption": 0.00011383432613620926,
+    "base_token_generation_latency_sync": 54.04956970214844,
+    "base_token_generation_latency_async": 54.09220550209284,
+    "base_token_generation_throughput_sync": 0.01850153489677552,
+    "base_token_generation_throughput_async": 0.018486951876297775,
+    "base_token_generation_CO2_emissions": null,
+    "base_token_generation_energy_consumption": null,
+    "base_inference_latency_sync": 53.23417663574219,
+    "base_inference_latency_async": 52.190279960632324,
+    "base_inference_throughput_sync": 0.018784924708097873,
+    "base_inference_throughput_async": 0.01916065598334231,
+    "base_inference_CO2_emissions": null,
+    "base_inference_energy_consumption": null,
     "smashed_current_gpu_type": "NVIDIA A100-PCIE-40GB",
     "smashed_current_gpu_total_memory": 40339.3125,
-    "smashed_token_generation_latency_sync": 49.12493553161621,
-    "smashed_token_generation_latency_async": 48.867019824683666,
-    "smashed_token_generation_throughput_sync": 0.02035626081089536,
-    "smashed_token_generation_throughput_async": 0.020463699312698435,
-    "smashed_token_generation_CO2_emissions": 3.728837750948082e-05,
-    "smashed_token_generation_energy_consumption": 0.009091379798953737,
-    "smashed_inference_latency_sync": 52.04664306640625,
-    "smashed_inference_latency_async": 49.13980960845947,
-    "smashed_inference_throughput_sync": 0.01921353503479756,
-    "smashed_inference_throughput_async": 0.02035009919590427,
-    "smashed_inference_CO2_emissions": 3.594376337961103e-05,
-    "smashed_inference_energy_consumption": 0.00010089579813269325
+    "smashed_token_generation_latency_sync": 40.57583160400391,
+    "smashed_token_generation_latency_async": 40.6801862642169,
+    "smashed_token_generation_throughput_sync": 0.024645212691126284,
+    "smashed_token_generation_throughput_async": 0.024581991672924564,
+    "smashed_token_generation_CO2_emissions": null,
+    "smashed_token_generation_energy_consumption": null,
+    "smashed_inference_latency_sync": 51.8927360534668,
+    "smashed_inference_latency_async": 41.21739864349365,
+    "smashed_inference_throughput_sync": 0.019270519846355125,
+    "smashed_inference_throughput_async": 0.02426159905552056,
+    "smashed_inference_CO2_emissions": null,
+    "smashed_inference_energy_consumption": null
 }

smash_config.json

@@ -2,19 +2,19 @@
     "api_key": null,
     "verify_url": "http://johnrachwan.pythonanywhere.com",
     "smash_config": {
-        "pruners": "[]",
+        "pruners": "None",
         "pruning_ratio": 0.0,
-        "factorizers": "[]",
+        "factorizers": "None",
         "quantizers": "['awq']",
         "weight_quantization_bits": 4,
-        "output_deviation": 0.01,
-        "compilers": "[]",
+        "output_deviation": 0.005,
+        "compilers": "None",
         "static_batch": true,
         "static_shape": true,
         "controlnet": "None",
         "unet_dim": 4,
         "device": "cuda",
-        "cache_dir": "/ceph/hdd/staff/charpent/.cache/modelshwuv6vps",
+        "cache_dir": "/ceph/hdd/staff/charpent/.cache/modelsa_klkits",
         "batch_size": 1,
         "model_name": "cognitivecomputations/dolphin-2.9-llama3-8b-256k",
         "task": "text_text_generation",