Upload folder using huggingface_hub

README.md

@@ -11,15 +11,6 @@ tags:
   - qwen3
 size_categories:
   - 10M<n<100M
-configs:
-  - config_name: trace
-    data_files:
-      - split: train
-        path: trace.jsonl
-  - config_name: qwen3-32b-buckets
-    data_files:
-      - split: train
-        path: qwen3-32b-buckets.jsonl
 ---
 
 # LLM Serving Trace
@@ -37,30 +28,51 @@ input-token buckets. It is packaged as newline-delimited JSON.
 
 ## Loading
 
+The files are distributed as raw newline-delimited JSON. The `model_parameters`
+field is heterogeneous across requests (mixed types and occasional extra keys),
+so the schema is not friendly to `datasets.load_dataset`'s automatic Arrow
+inference. Download the files with `huggingface_hub` and read them directly.
+
 ```python
-from datasets import load_dataset
+import json
+from pathlib import Path
+
+from huggingface_hub import snapshot_download
 
-trace = load_dataset("eth-easl/swissai-serving-trace", "trace", split="train")
-qwen_buckets = load_dataset(
-    "eth-easl/swissai-serving-trace",
-    "qwen3-32b-buckets",
-    split="train",
+local_dir = snapshot_download(
+    repo_id="eth-easl/swissai-serving-trace",
+    repo_type="dataset",
+    allow_patterns=["trace.jsonl"],  # or ["qwen3-32b-buckets.jsonl"], or both
 )
+
+with (Path(local_dir) / "trace.jsonl").open("r", encoding="utf-8") as handle:
+    for line in handle:
+        record = json.loads(line)
+        # process record
 ```
 
-For low-memory iteration, use streaming:
+To grab a single file without materialising the whole snapshot tree, use
+`hf_hub_download`:
 
 ```python
-from datasets import load_dataset
+from huggingface_hub import hf_hub_download
 
-rows = load_dataset(
-    "eth-easl/swissai-serving-trace",
-    "qwen3-32b-buckets",
-    split="train",
-    streaming=True,
+path = hf_hub_download(
+    repo_id="eth-easl/swissai-serving-trace",
+    filename="qwen3-32b-buckets.jsonl",
+    repo_type="dataset",
 )
+```
+
+If you prefer a dataframe and have enough memory, pandas can read the JSONL
+directly (`trace.jsonl` is ~7 GB, so consider `chunksize=`):
+
+```python
+import pandas as pd
 
-first_row = next(iter(rows))
+frames = pd.read_json(path, lines=True, chunksize=100_000)
+for chunk in frames:
+    ...  # process chunk
 ```
 
 ## Schema
@@ -80,16 +92,21 @@ Each row contains:
 | `reported_token_input` | int64 | Reported input-token count, or `-1` when unavailable. |
 | `reported_token_output` | int64 | Reported output-token count, or `-1` when unavailable. |
 
-`model_parameters` contains:
+`model_parameters` is the raw user-supplied generation config and is
+deliberately not normalised. Most rows contain the keys below, but values are
+mixed JSON types because they reflect what the client sent:
 
-| Field | Type |
+| Field | Observed JSON types |
 | --- | --- |
-| `temperature` | float64 |
-| `max_tokens` | string |
-| `top_p` | float64 |
-| `frequency_penalty` | float64 |
-| `presence_penalty` | float64 |
-| `seed` | int64 |
+| `temperature` | number, `"null"` string, missing |
+| `max_tokens` | number, string, `null` |
+| `top_p` | number, `"null"` string, missing |
+| `frequency_penalty` | number, string, list, `null`, missing |
+| `presence_penalty` | number, string, bool, `null`, missing |
+| `seed` | integer, missing |
+
+A small number of rows also carry an extra `n` integer key. Coerce these to
+your preferred schema at read time.
 
 ### `qwen3-32b-buckets`