Merge 73aa52a into 1261fe8

Author: DaltheCow

src/guidellm/presentation/data_models.py

@@ -1,14 +1,21 @@
+import httpx
 import random
 from collections import defaultdict
 from math import ceil
-from typing import TYPE_CHECKING, Optional, Union
+from typing import TYPE_CHECKING, Any, Optional, Union
 
 from pydantic import BaseModel, computed_field
 
 if TYPE_CHECKING:
     from guidellm.benchmark.benchmark import GenerativeBenchmark
 
+from guidellm.dataset.file import FileDatasetCreator
+from guidellm.dataset.hf_datasets import HFDatasetsCreator
+from guidellm.dataset.in_memory import InMemoryDatasetCreator
+from guidellm.dataset.synthetic import SyntheticDatasetConfig, SyntheticDatasetCreator
 from guidellm.objects.statistics import DistributionSummary
+from guidellm.preprocess.dataset import TokensConfig
+
 
 
 class Bucket(BaseModel):
@@ -58,6 +65,38 @@ class Model(BaseModel):
 class Dataset(BaseModel):
     name: str
 
+    @classmethod
+    def from_data(cls, request_loader: Any):
+        creators = [
+            InMemoryDatasetCreator,
+            SyntheticDatasetCreator,
+            FileDatasetCreator,
+            HFDatasetsCreator,
+        ]
+        dataset_name = ""
+        data = request_loader.data
+        data_args = request_loader.data_args
+        processor = request_loader.processor
+        processor_args = request_loader.processor_args
+        
+        for creator in creators:
+            if creator.is_supported(data, None):
+                random_seed = 42
+                dataset = creator.handle_create(data, data_args, processor, processor_args, random_seed)
+                dataset_name = creator.extract_dataset_name(dataset)
+                if dataset_name is None or dataset_name == "":
+                    if creator == SyntheticDatasetCreator:
+                        data_dict = SyntheticDatasetConfig.parse_str(data)
+                        dataset_name = data_dict.source
+                    if creator == FileDatasetCreator or isinstance(creator, HFDatasetsCreator):
+                        dataset_name = data
+                    if creator == InMemoryDatasetCreator:
+                        dataset_name = "In-memory"
+                break
+        return cls(
+            name=dataset_name or ""
+        )
+
 
 class RunInfo(BaseModel):
     model: Model
@@ -71,13 +110,16 @@ def from_benchmarks(cls, benchmarks: list["GenerativeBenchmark"]):
         timestamp = max(
             bm.run_stats.start_time for bm in benchmarks if bm.start_time is not None
         )
+        response = httpx.get(f"https://huggingface.co/api/models/{model}")
+        modelJson = response.json()
+            
         return cls(
-            model=Model(name=model, size=0),
+            model=Model(name=model, size=modelJson.get("usedStorage", 0)),
             task="N/A",
             timestamp=timestamp,
-            dataset=Dataset(name="N/A"),
+            dataset=Dataset.from_data(benchmarks[0].request_loader),
         )
-
+    
 
 class Distribution(BaseModel):
     statistics: Optional[DistributionSummary] = None

src/ui/lib/components/PageHeader/PageHeader.component.tsx

@@ -2,12 +2,14 @@
 import { Box, Typography } from '@mui/material';
 
 import { useGetRunInfoQuery } from '../../store/slices/runInfo';
-import { formateDate } from '../../utils/helpers';
+import { formateDate, getFileSize } from '../../utils/helpers';
 import { SpecBadge } from '../SpecBadge';
 import { HeaderCell, HeaderWrapper } from './PageHeader.styles';
 
 export const Component = () => {
   const { data } = useGetRunInfoQuery();
+  const modelSize = getFileSize(data?.model?.size || 0);
+
   return (
     <Box py={2}>
       <Typography variant="subtitle2" color="surface.onSurfaceAccent">
@@ -24,11 +26,24 @@ export const Component = () => {
             variant="metric2"
             withTooltip
           />
+          <SpecBadge
+            label="Model size"
+            value={data?.model?.size ? `${modelSize?.size} ${modelSize?.units}` : '0B'}
+            variant="body1"
+          />
+        </HeaderCell>
+        <HeaderCell item xs={5} withDivider>
+          <SpecBadge
+            label="Dataset"
+            value={data?.dataset?.name || 'N/A'}
+            variant="caption"
+            withTooltip
+          />
         </HeaderCell>
         <HeaderCell item xs={2} sx={{ paddingRight: 0 }}>
           <SpecBadge
             label="Time Stamp"
-            value={data?.timestamp ? formateDate(data?.timestamp) : 'n/a'}
+            value={data?.timestamp ? formateDate(data?.timestamp) : 'N/A'}
             variant="caption"
           />
         </HeaderCell>