Sync from GitHub via hub-sync

lighton-ocr2-saturate.py

@@ -1,7 +1,7 @@
 # /// script
 # requires-python = ">=3.11"
 # dependencies = [
-#     "saturate[hf]>=0.1.1",
+#     "saturate[hf]==0.3.0",
 #     "pillow>=10",
 # ]
 #
@@ -135,6 +135,7 @@ def main():
                     help="Re-admit rows whose only record is an error row")
     args = ap.parse_args()
 
+    import pyarrow as pa
     from saturate import Auto, Engine, dataset_rows, pump
 
     target_size = 0 if args.no_resize else args.target_size
@@ -155,6 +156,13 @@ def main():
             "max_tokens": args.max_tokens,
         }
 
+    # Declared, so a part holding only error rows has the same columns as the rest
+    # (datasets.load_dataset reads every part with one schema).
+    schema = pa.schema([
+        ("id", pa.string()), ("error", pa.string()), ("markdown", pa.string()),
+        ("model", pa.string()), ("prompt_tokens", pa.int64()), ("completion_tokens", pa.int64()),
+    ])  # fmt: skip
+
     def parse(row, body):
         usage = body.get("usage") or {}
         return {
@@ -169,7 +177,7 @@ def main():
     with Engine(SERVING["model"], engine="vllm", extra_args=extra) as endpoint:
         stats = pump(rows, to_request, parse, endpoint, output,
                      window=Auto(initial=8, max_limit=48),
-                     retry_errors=args.retry_errors)
+                     retry_errors=args.retry_errors, schema=schema)
 
     print(f"https://huggingface.co/datasets/{args.output_dataset} "
           f"({stats.rows_processed} ok, {stats.rows_failed} error rows)", file=sys.stderr)

ovis-ocr2-saturate.py

@@ -1,7 +1,7 @@
 # /// script
 # requires-python = ">=3.11"
 # dependencies = [
-#     "saturate[hf]>=0.1.1",
+#     "saturate[hf]==0.3.0",
 #     "pillow>=10",
 # ]
 #
@@ -196,6 +196,7 @@ def main():
                     help="Re-admit rows whose only record is an error row")
     args = ap.parse_args()
 
+    import pyarrow as pa
     from saturate import Auto, Engine, dataset_rows, pump
 
     rows = dataset_rows(
@@ -216,6 +217,13 @@ def main():
             "chat_template_kwargs": {"enable_thinking": False},
         }
 
+    # Declared, so a part holding only error rows has the same columns as the rest
+    # (datasets.load_dataset reads every part with one schema).
+    schema = pa.schema([
+        ("id", pa.string()), ("error", pa.string()), ("markdown", pa.string()),
+        ("model", pa.string()), ("prompt_tokens", pa.int64()), ("completion_tokens", pa.int64()),
+    ])  # fmt: skip
+
     def parse(row, body):
         text = body["choices"][0]["message"]["content"].strip()
         if not args.keep_image_tags:
@@ -234,7 +242,7 @@ def main():
     with Engine(SERVING["model"], engine="vllm", extra_args=extra) as endpoint:
         stats = pump(rows, to_request, parse, endpoint, output,
                      window=Auto(initial=8, max_limit=48),
-                     retry_errors=args.retry_errors)
+                     retry_errors=args.retry_errors, schema=schema)
 
     print(f"https://huggingface.co/datasets/{args.output_dataset} "
           f"({stats.rows_processed} ok, {stats.rows_failed} error rows)", file=sys.stderr)

unlimited-ocr2-live.py

@@ -1,7 +1,7 @@
 # /// script
 # requires-python = ">=3.11"
 # dependencies = [
-#     "saturate[hf]==0.2.0",
+#     "saturate[hf]==0.3.0",
 #     "pillow>=10",
 # ]
 #
@@ -691,7 +691,6 @@ def main():
     args = ap.parse_args()
 
     import fsspec
-    import saturate.transport
     from saturate import Auto, Engine, pump
 
     rows = list(list_windows(args.input, args.pages_per_request, args.max_pages))
@@ -736,9 +735,8 @@ def main():
     is_repo_id = re.fullmatch(r"[\w-][\w.-]*/[\w.-]*[\w-]", args.output) and not os.path.exists(args.output)
     output = f"hf://datasets/{args.output}/data" if is_repo_id else args.output
 
-    # One request is minutes of decoding. saturate 0.2.0 caps a request's retries at
-    # 300 s in total, which a retried 16-page request never fits in; match the read timeout.
-    saturate.transport.RETRY_BUDGET_S = 3600.0
+    # One request is minutes of decoding: read_timeout=3600 below, and a row's retries get
+    # 2 x read_timeout in total (saturate's default retry budget).
     if live is not None:
         live.phase("Installing Baidu's SGLang build")
     cmd = start_server_cmd()

unlimited-ocr2-saturate.py

@@ -1,7 +1,7 @@
 # /// script
 # requires-python = ">=3.11"
 # dependencies = [
-#     "saturate[hf]==0.2.0",
+#     "saturate[hf]==0.3.0",
 #     "pillow>=10",
 #     "pymupdf>=1.24",
 # ]
@@ -527,14 +527,8 @@ def main():
     # headroom (a100-large: 8 x 32-page windows used ~25% of KV) and flush often,
     # since each row is expensive to redo.
     window = Auto(initial=32, max_limit=64)
-    # saturate caps a row's retries at RETRY_BUDGET_S (300 s, counted from the first
-    # attempt), which is shorter than one legitimate multi-page generation here, so a
-    # transient first-attempt failure always ended in a TimeoutError row. Match the
-    # read timeout until saturate exposes this as a pump() argument.
-    import saturate.transport
-
-    saturate.transport.RETRY_BUDGET_S = 3600.0
-
+    # read_timeout=3600: one window is minutes of decode. A row's retries then get
+    # 2 x read_timeout in total (saturate's default retry budget).
     def run_passes(endpoint, headers=None):
         common = {"retry_errors": args.retry_errors, "headers": headers,
                   "prepare_workers": args.prepare_workers, "read_timeout": 3600}

unlimited-ocr2.py

@@ -1,7 +1,7 @@
 # /// script
 # requires-python = ">=3.11"
 # dependencies = [
-#     "saturate[hf]==0.2.0",
+#     "saturate[hf]==0.3.0",
 #     "pillow>=10",
 # ]
 #
@@ -279,7 +279,6 @@ def main():
     SERVING["env"]["SGLANG_UOCR_MAX_PAGE_TOKENS"] = str(args.max_page_tokens)
 
     import fsspec
-    import saturate.transport
     from saturate import Auto, Engine, pump
 
     rows = list(list_windows(args.input, args.pages_per_request, args.max_pages))
@@ -335,9 +334,8 @@ def main():
     is_repo_id = re.fullmatch(r"[\w-][\w.-]*/[\w.-]*[\w-]", args.output) and not os.path.exists(args.output)
     output = f"hf://datasets/{args.output}/data" if is_repo_id else args.output
 
-    # One request is minutes of decoding. saturate 0.2.0 caps a request's retries at
-    # 300 s in total, which a retried 16-page request never fits in; match the read timeout.
-    saturate.transport.RETRY_BUDGET_S = 3600.0
+    # One request is minutes of decoding: read_timeout=3600 below, and a row's retries get
+    # 2 x read_timeout in total (saturate's default retry budget).
     with Engine(SERVING["model"], engine="sglang", cmd=start_server_cmd()) as endpoint:
         stats = pump(rows, to_request, parse, endpoint, output,
                      window=Auto(initial=32, max_limit=64),  # long requests: start wide