Spaces:
Running
Running
Julien Simon
commited on
Commit
•
9f46468
1
Parent(s):
0c0f086
Fix N/A values
Browse files- app.py +10 -5
- results_llama_spark.py +10 -0
app.py
CHANGED
@@ -9,7 +9,7 @@ import re
|
|
9 |
import gradio as gr
|
10 |
import pandas as pd
|
11 |
|
12 |
-
from results import results
|
13 |
|
14 |
logging.basicConfig(level=logging.DEBUG)
|
15 |
|
@@ -107,14 +107,19 @@ def display_results(model_name):
|
|
107 |
cloud = config.get("cloud", "N/A")
|
108 |
instance_type = config.get("instanceType", "N/A")
|
109 |
|
|
|
|
|
|
|
|
|
|
|
110 |
if "configurations" in config:
|
111 |
for nested_config in config["configurations"]:
|
112 |
data.append(
|
113 |
{
|
114 |
"Cloud": cloud,
|
115 |
"Instance Type": instance_type,
|
116 |
-
"GPU":
|
117 |
-
"GPU RAM":
|
118 |
"Status": nested_config.get("status", "N/A"),
|
119 |
"Quantization": nested_config.get(
|
120 |
"quantization", "N/A"
|
@@ -134,8 +139,8 @@ def display_results(model_name):
|
|
134 |
{
|
135 |
"Cloud": cloud,
|
136 |
"Instance Type": instance_type,
|
137 |
-
"GPU":
|
138 |
-
"GPU RAM":
|
139 |
"Status": config.get("status", "N/A"),
|
140 |
"Quantization": config.get("quantization", "N/A"),
|
141 |
"Container": config.get(
|
|
|
9 |
import gradio as gr
|
10 |
import pandas as pd
|
11 |
|
12 |
+
from results import instance_type_mappings, results
|
13 |
|
14 |
logging.basicConfig(level=logging.DEBUG)
|
15 |
|
|
|
107 |
cloud = config.get("cloud", "N/A")
|
108 |
instance_type = config.get("instanceType", "N/A")
|
109 |
|
110 |
+
# Fetch GPU and GPU RAM information from instance_type_mappings
|
111 |
+
instance_info = instance_type_mappings.get(instance_type, {})
|
112 |
+
gpu = instance_info.get("gpu", "N/A")
|
113 |
+
gpu_ram = instance_info.get("gpuRAM", "N/A")
|
114 |
+
|
115 |
if "configurations" in config:
|
116 |
for nested_config in config["configurations"]:
|
117 |
data.append(
|
118 |
{
|
119 |
"Cloud": cloud,
|
120 |
"Instance Type": instance_type,
|
121 |
+
"GPU": gpu,
|
122 |
+
"GPU RAM": gpu_ram,
|
123 |
"Status": nested_config.get("status", "N/A"),
|
124 |
"Quantization": nested_config.get(
|
125 |
"quantization", "N/A"
|
|
|
139 |
{
|
140 |
"Cloud": cloud,
|
141 |
"Instance Type": instance_type,
|
142 |
+
"GPU": gpu,
|
143 |
+
"GPU RAM": gpu_ram,
|
144 |
"Status": config.get("status", "N/A"),
|
145 |
"Quantization": config.get("quantization", "N/A"),
|
146 |
"Container": config.get(
|
results_llama_spark.py
CHANGED
@@ -54,6 +54,16 @@ results_llama_spark = {
|
|
54 |
{"container": "vLLM 0.5.5", "tokensPerSecond": "43.4"},
|
55 |
],
|
56 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
{
|
58 |
"instanceType": "p4d.24xlarge",
|
59 |
"quantization": "none",
|
|
|
54 |
{"container": "vLLM 0.5.5", "tokensPerSecond": "43.4"},
|
55 |
],
|
56 |
},
|
57 |
+
{
|
58 |
+
"instanceType": "g6e.12xlarge",
|
59 |
+
"quantization": "none",
|
60 |
+
"status": "OK",
|
61 |
+
"configurations": [
|
62 |
+
{"container": "TGI 2.2.0", "tokensPerSecond": "112"},
|
63 |
+
{"container": "SGLang 0.2.13", "tokensPerSecond": "123"},
|
64 |
+
{"container": "vLLM 0.5.5", "tokensPerSecond": "106"},
|
65 |
+
],
|
66 |
+
},
|
67 |
{
|
68 |
"instanceType": "p4d.24xlarge",
|
69 |
"quantization": "none",
|