begumcig commited on
Commit
48cf990
·
verified ·
1 Parent(s): 0eedb54

Upload folder using huggingface_hub (#2)

Browse files

- 7d21961d7f6f19ef5e680dfe7ffb787b50a07f8d3a95b6a3941ca100f1888a01 (1dbef9fa1ac6d6c32a4a9bf0eabce5ed7e476565)

Files changed (3) hide show
  1. base_results.json +19 -0
  2. plots.png +0 -0
  3. smashed_results.json +19 -0
base_results.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "current_gpu_type": "Tesla T4",
3
+ "current_gpu_total_memory": 15095.0625,
4
+ "perplexity": 2.2062017917633057,
5
+ "memory_inference_first": 7278.0,
6
+ "memory_inference": 7278.0,
7
+ "token_generation_latency_sync": 36.65739364624024,
8
+ "token_generation_latency_async": 36.3558666780591,
9
+ "token_generation_throughput_sync": 0.02727962630541697,
10
+ "token_generation_throughput_async": 0.027505877080452153,
11
+ "token_generation_CO2_emissions": 0.00018283805669861312,
12
+ "token_generation_energy_consumption": 0.002242028576060243,
13
+ "inference_latency_sync": 1057.4123092651366,
14
+ "inference_latency_async": 136.27676963806152,
15
+ "inference_throughput_sync": 0.000945704897926679,
16
+ "inference_throughput_async": 0.007338007810545461,
17
+ "inference_CO2_emissions": 0.00018305297379900547,
18
+ "inference_energy_consumption": 0.0006297497502826921
19
+ }
plots.png ADDED
smashed_results.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "current_gpu_type": "Tesla T4",
3
+ "current_gpu_total_memory": 15095.0625,
4
+ "perplexity": 2.548238754272461,
5
+ "memory_inference_first": 1458.0,
6
+ "memory_inference": 1458.0,
7
+ "token_generation_latency_sync": 57.3639274597168,
8
+ "token_generation_latency_async": 57.42117892950773,
9
+ "token_generation_throughput_sync": 0.01743255812988466,
10
+ "token_generation_throughput_async": 0.017415177093936633,
11
+ "token_generation_CO2_emissions": 0.00017302643152991706,
12
+ "token_generation_energy_consumption": 0.003290108461267099,
13
+ "inference_latency_sync": 964.0653579711914,
14
+ "inference_latency_async": 333.02977085113525,
15
+ "inference_throughput_sync": 0.0010372740724803456,
16
+ "inference_throughput_async": 0.0030027345526625645,
17
+ "inference_CO2_emissions": 0.0001732807294008593,
18
+ "inference_energy_consumption": 0.0005966248434902247
19
+ }