sharpenb commited on
Commit
ebbf76a
1 Parent(s): f13c7f4

1ff3d25cf63eeb7d4db79ea99406cccfb60dae0b5906cd6b07533efe2eb32261

Browse files
Files changed (4) hide show
  1. config.json +1 -1
  2. plots.png +0 -0
  3. results.json +24 -24
  4. smash_config.json +5 -5
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/tmp/tmpcmzh3zye",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
 
1
  {
2
+ "_name_or_path": "/tmp/tmpu4qxllsm",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
plots.png CHANGED
results.json CHANGED
@@ -1,30 +1,30 @@
1
  {
2
  "base_current_gpu_type": "NVIDIA A100-PCIE-40GB",
3
  "base_current_gpu_total_memory": 40339.3125,
4
- "base_token_generation_latency_sync": 76.9214370727539,
5
- "base_token_generation_latency_async": 76.84411387890577,
6
- "base_token_generation_throughput_sync": 0.01300027714061269,
7
- "base_token_generation_throughput_async": 0.01301335846719298,
8
- "base_token_generation_CO2_emissions": 4.107156647942229e-05,
9
- "base_token_generation_energy_consumption": 0.013795784438629225,
10
- "base_inference_latency_sync": 75.23112869262695,
11
- "base_inference_latency_async": 74.10402297973633,
12
- "base_inference_throughput_sync": 0.01329237002525532,
13
- "base_inference_throughput_async": 0.013494544017852432,
14
- "base_inference_CO2_emissions": 4.120221930394127e-05,
15
- "base_inference_energy_consumption": 0.00011383432613620926,
16
  "smashed_current_gpu_type": "NVIDIA A100-PCIE-40GB",
17
  "smashed_current_gpu_total_memory": 40339.3125,
18
- "smashed_token_generation_latency_sync": 49.12493553161621,
19
- "smashed_token_generation_latency_async": 48.867019824683666,
20
- "smashed_token_generation_throughput_sync": 0.02035626081089536,
21
- "smashed_token_generation_throughput_async": 0.020463699312698435,
22
- "smashed_token_generation_CO2_emissions": 3.728837750948082e-05,
23
- "smashed_token_generation_energy_consumption": 0.009091379798953737,
24
- "smashed_inference_latency_sync": 52.04664306640625,
25
- "smashed_inference_latency_async": 49.13980960845947,
26
- "smashed_inference_throughput_sync": 0.01921353503479756,
27
- "smashed_inference_throughput_async": 0.02035009919590427,
28
- "smashed_inference_CO2_emissions": 3.594376337961103e-05,
29
- "smashed_inference_energy_consumption": 0.00010089579813269325
30
  }
 
1
  {
2
  "base_current_gpu_type": "NVIDIA A100-PCIE-40GB",
3
  "base_current_gpu_total_memory": 40339.3125,
4
+ "base_token_generation_latency_sync": 54.04956970214844,
5
+ "base_token_generation_latency_async": 54.09220550209284,
6
+ "base_token_generation_throughput_sync": 0.01850153489677552,
7
+ "base_token_generation_throughput_async": 0.018486951876297775,
8
+ "base_token_generation_CO2_emissions": null,
9
+ "base_token_generation_energy_consumption": null,
10
+ "base_inference_latency_sync": 53.23417663574219,
11
+ "base_inference_latency_async": 52.190279960632324,
12
+ "base_inference_throughput_sync": 0.018784924708097873,
13
+ "base_inference_throughput_async": 0.01916065598334231,
14
+ "base_inference_CO2_emissions": null,
15
+ "base_inference_energy_consumption": null,
16
  "smashed_current_gpu_type": "NVIDIA A100-PCIE-40GB",
17
  "smashed_current_gpu_total_memory": 40339.3125,
18
+ "smashed_token_generation_latency_sync": 40.57583160400391,
19
+ "smashed_token_generation_latency_async": 40.6801862642169,
20
+ "smashed_token_generation_throughput_sync": 0.024645212691126284,
21
+ "smashed_token_generation_throughput_async": 0.024581991672924564,
22
+ "smashed_token_generation_CO2_emissions": null,
23
+ "smashed_token_generation_energy_consumption": null,
24
+ "smashed_inference_latency_sync": 51.8927360534668,
25
+ "smashed_inference_latency_async": 41.21739864349365,
26
+ "smashed_inference_throughput_sync": 0.019270519846355125,
27
+ "smashed_inference_throughput_async": 0.02426159905552056,
28
+ "smashed_inference_CO2_emissions": null,
29
+ "smashed_inference_energy_consumption": null
30
  }
smash_config.json CHANGED
@@ -2,19 +2,19 @@
2
  "api_key": null,
3
  "verify_url": "http://johnrachwan.pythonanywhere.com",
4
  "smash_config": {
5
- "pruners": "[]",
6
  "pruning_ratio": 0.0,
7
- "factorizers": "[]",
8
  "quantizers": "['awq']",
9
  "weight_quantization_bits": 4,
10
- "output_deviation": 0.01,
11
- "compilers": "[]",
12
  "static_batch": true,
13
  "static_shape": true,
14
  "controlnet": "None",
15
  "unet_dim": 4,
16
  "device": "cuda",
17
- "cache_dir": "/ceph/hdd/staff/charpent/.cache/modelshwuv6vps",
18
  "batch_size": 1,
19
  "model_name": "cognitivecomputations/dolphin-2.9-llama3-8b-256k",
20
  "task": "text_text_generation",
 
2
  "api_key": null,
3
  "verify_url": "http://johnrachwan.pythonanywhere.com",
4
  "smash_config": {
5
+ "pruners": "None",
6
  "pruning_ratio": 0.0,
7
+ "factorizers": "None",
8
  "quantizers": "['awq']",
9
  "weight_quantization_bits": 4,
10
+ "output_deviation": 0.005,
11
+ "compilers": "None",
12
  "static_batch": true,
13
  "static_shape": true,
14
  "controlnet": "None",
15
  "unet_dim": 4,
16
  "device": "cuda",
17
+ "cache_dir": "/ceph/hdd/staff/charpent/.cache/modelsa_klkits",
18
  "batch_size": 1,
19
  "model_name": "cognitivecomputations/dolphin-2.9-llama3-8b-256k",
20
  "task": "text_text_generation",