Vadim Kulibaba recoilme

Model	AGIEval	GPT4All	TruthfulQA	Bigbench
recoilme-gemma-2-9B-v0.4	45.94	Error: File does not exist	59.03	Error: File does not exist

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	26.77	±	2.78
		acc_norm	25.98	±	2.76
agieval_logiqa_en	0	acc	41.32	±	1.93

Model	ARC	HellaSwag	MMLU	TruthfulQA	Winogrande	GSM8K
recoilme-gemma-2-9B-v0.4	Error: File does not exist	Error: File does not exist	Error: File does not exist	Error: File does not exist	Error: File does not exist	Error: File does not exist

Average: Error: File does not exist%

Model	AGIEval	GPT4All	TruthfulQA	Bigbench
recoilme-gemma-2-9B-v0.3	40.71	Error: File does not exist	58.62	Error: File does not exist

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	22.44	±	2.62
		acc_norm	24.41	±	2.70
agieval_logiqa_en	0	acc	36.56	±	1.89

Model	AGIEval	GPT4All	TruthfulQA	Bigbench
Gemma-2-Ataraxy-Gemmasutra-9B-slerp	40.91	Error: File does not exist	60.1	Error: File does not exist

Task	Version	Metric	Value		Stderr
agieval_aqua_rat	0	acc	21.26	±	2.57
		acc_norm	23.23	±	2.65
agieval_logiqa_en	0	acc	38.56	±	1.91

	import os
	import shutil
	import zlib
	import zipfile

	import cv2
	import numpy as np
	from PIL import Image

	def downscale_image_by(image, max_size,x=64):

	from diffusers import AutoencoderTiny, AutoencoderKL
	import torch
	import torchvision.transforms.functional as TF
	from PIL import Image

	#taesd3 = AutoencoderTiny.from_pretrained(
	# "madebyollin/taesd3", torch_dtype=torch.float16
	#).to("cuda")
	sd3_vae = AutoencoderKL.from_pretrained(
	"AuraDiffusion/16ch-vae",

	from http.server import BaseHTTPRequestHandler, HTTPServer
	import io
	import torch, gc
	import base64
	import json
	import numpy as np
	from PIL import Image
	from datetime import datetime
	import time

	import cv2
	import glob
	import numpy as np
	import os
	import torch

	#download: https://github.com/xinntao/ESRGAN/blob/master/RRDBNet_arch.py
	import RRDBNet_arch as arch

	model_path = '/home/recoilme/forge/models/ESRGAN/ESRGAN_4x.pth'