ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge pull request #3869 from hnxnq7/fix-kaggle-telemetry-detection

Browse source 
Fix Kaggle telemetry misclassification when COLAB_ keys exist
This commit is contained in:
Daniel Han 2026-01-08 17:29:23 -08:00 • committed by GitHub
commit 0e54b817af
1 changed files with 58 additions and 45 deletions
Download patch file Download diff file Expand all files Collapse all files

103
unsloth/models/_utils.py
View file

@ -1106,53 +1106,66 @@ def _get_statistics(statistics = None, force_download = True):
global USE_MODELSCOPE
USE_MODELSCOPE = os.environ.get("UNSLOTH_USE_MODELSCOPE", "0") == "1"
if statistics is not None:
pass
elif "\nCOLAB_" in keynames and n_cpus == 1:
statistics = "colab"
elif "\nCOLAB_" in keynames:
statistics = "colabpro"
elif "\nKAGGLE_" in keynames:
statistics = "kaggle"
elif "\nRUNPOD_" in keynames:
statistics = "runpod"
elif "\nAWS_" in keynames:
statistics = "aws"
elif "\nAZURE_" in keynames:
statistics = "azure"
# elif "\nK_" in keynames or "\nFUNCTION_" in keynames: statistics = "gcp"
elif "\nINVOCATION_ID" in keynames:
statistics = "lambda"
# else: statistics = "other"
else:
def try_vllm_check():
vendor_files = (
"/sys/class/dmi/id/product_version",
"/sys/class/dmi/id/bios_vendor",
"/sys/class/dmi/id/product_name",
"/sys/class/dmi/id/chassis_asset_tag",
"/sys/class/dmi/id/sys_vendor",
)
if statistics is None:
# Prefer filesystem markers (harder to misidentify) before env-key matching
try:
from pathlib import Path
for vendor_file in vendor_files:
path = Path(vendor_file)
if path.is_file():
file_content = path.read_text().lower()
if "amazon" in file_content:
return "aws"
elif "microsoft corporation" in file_content:
return "azure"
elif "google" in file_content:
return "gcp"
return "other"
if Path("/kaggle/working").exists():
statistics = "kaggle"
elif Path("/content").exists() and Path("/opt/colab").exists():
statistics = "colab" if n_cpus == 1 else "colabpro"
elif Path("/runpod-volume").exists():
statistics = "runpod"
except Exception:
pass
# Fallback to env-key detection
if statistics is None:
if "\nKAGGLE_" in keynames:
statistics = "kaggle"
elif "\nCOLAB_" in keynames and n_cpus == 1:
statistics = "colab"
elif "\nCOLAB_" in keynames:
statistics = "colabpro"
elif "\nRUNPOD_" in keynames:
statistics = "runpod"
elif "\nAWS_" in keynames:
statistics = "aws"
elif "\nAZURE_" in keynames:
statistics = "azure"
# elif "\nK_" in keynames or "\nFUNCTION_" in keynames: statistics = "gcp"
elif "\nINVOCATION_ID" in keynames:
statistics = "lambda"
# else: statistics = "other"
else:
def try_vllm_check():
vendor_files = (
"/sys/class/dmi/id/product_version",
"/sys/class/dmi/id/bios_vendor",
"/sys/class/dmi/id/product_name",
"/sys/class/dmi/id/chassis_asset_tag",
"/sys/class/dmi/id/sys_vendor",
)
for vendor_file in vendor_files:
path = Path(vendor_file)
if path.is_file():
file_content = path.read_text().lower()
if "amazon" in file_content:
return "aws"
elif "microsoft corporation" in file_content:
return "azure"
elif "google" in file_content:
return "gcp"
return "other"
try:
statistics = try_vllm_check()
except Exception:
statistics = "other"
pass
try:
statistics = try_vllm_check()
except:
statistics = "other"
if statistics is not None:
import tempfile
from huggingface_hub import snapshot_download
@ -1184,7 +1197,7 @@ def _get_statistics(statistics = None, force_download = True):
"model = FastLanguageModel.from_pretrained('unsloth/gpt-oss-20b')\n"
"```"
)
except:
except Exception:
# Try no time limit check
stats_check()