!git clone https://github.com/OpenAccess-AI-Collective/axolotl.git

Cloning into 'axolotl'...
remote: Enumerating objects: 2614, done.
remote: Counting objects: 100% (890/890), done.
remote: Compressing objects: 100% (195/195), done.
remote: Total 2614 (delta 791), reused 721 (delta 685), pack-reused 1724
Receiving objects: 100% (2614/2614), 1.33 MiB | 9.54 MiB/s, done.
Resolving deltas: 100% (1627/1627), done.

%cd axolotl

/content/axolotl

!pip install -e .

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Obtaining file:///content/axolotl
  Preparing metadata (setup.py) ... done
Collecting transformers@ git+https://github.com/huggingface/transformers.git (from axolotl==0.1)
  Cloning https://github.com/huggingface/transformers.git to /tmp/pip-install-esok7o5q/transformers_6129b103593a488c99b68220b5a40d08
  Running command git clone --filter=blob:none --quiet https://github.com/huggingface/transformers.git /tmp/pip-install-esok7o5q/transformers_6129b103593a488c99b68220b5a40d08
  Resolved https://github.com/huggingface/transformers.git to commit 8f093fb799246f7dd9104ff44728da0c53a9f67a
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... done
Collecting bitsandbytes>=0.39.0 (from axolotl==0.1)
  Downloading bitsandbytes-0.39.0-py3-none-any.whl (92.2 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 92.2/92.2 MB 10.6 MB/s eta 0:00:00
Collecting accelerate (from axolotl==0.1)
  Downloading accelerate-0.20.3-py3-none-any.whl (227 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 227.6/227.6 kB 26.1 MB/s eta 0:00:00
Collecting addict (from axolotl==0.1)
  Downloading addict-2.4.0-py3-none-any.whl (3.8 kB)
Collecting fire (from axolotl==0.1)
  Downloading fire-0.5.0.tar.gz (88 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 88.3/88.3 kB 12.1 MB/s eta 0:00:00
  Preparing metadata (setup.py) ... done
Requirement already satisfied: PyYAML==6.0 in /usr/local/lib/python3.10/dist-packages (from axolotl==0.1) (6.0)
Collecting datasets (from axolotl==0.1)
  Downloading datasets-2.12.0-py3-none-any.whl (474 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 474.6/474.6 kB 52.0 MB/s eta 0:00:00
Collecting sentencepiece (from axolotl==0.1)
  Downloading sentencepiece-0.1.99-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.3 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.3/1.3 MB 80.8 MB/s eta 0:00:00
Collecting wandb (from axolotl==0.1)
  Downloading wandb-0.15.4-py3-none-any.whl (2.1 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 2.1/2.1 MB 101.9 MB/s eta 0:00:00
Collecting einops (from axolotl==0.1)
  Downloading einops-0.6.1-py3-none-any.whl (42 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 42.2/42.2 kB 6.1 MB/s eta 0:00:00
Collecting xformers (from axolotl==0.1)
  Downloading xformers-0.0.20-cp310-cp310-manylinux2014_x86_64.whl (109.1 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 109.1/109.1 MB 8.6 MB/s eta 0:00:00
Collecting bert-score==0.3.13 (from axolotl==0.1)
  Downloading bert_score-0.3.13-py3-none-any.whl (61 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 61.1/61.1 kB 8.2 MB/s eta 0:00:00
Collecting evaluate==0.4.0 (from axolotl==0.1)
  Downloading evaluate-0.4.0-py3-none-any.whl (81 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 81.4/81.4 kB 11.0 MB/s eta 0:00:00
Collecting rouge-score==0.1.2 (from axolotl==0.1)
  Downloading rouge_score-0.1.2.tar.gz (17 kB)
  Preparing metadata (setup.py) ... done
Requirement already satisfied: scipy in /usr/local/lib/python3.10/dist-packages (from axolotl==0.1) (1.10.1)
Requirement already satisfied: scikit-learn==1.2.2 in /usr/local/lib/python3.10/dist-packages (from axolotl==0.1) (1.2.2)
Requirement already satisfied: torch>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (2.0.1+cu118)
Requirement already satisfied: pandas>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (1.5.3)
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (1.22.4)
Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (2.27.1)
Requirement already satisfied: tqdm>=4.31.1 in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (4.65.0)
Requirement already satisfied: matplotlib in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (3.7.1)
Requirement already satisfied: packaging>=20.9 in /usr/local/lib/python3.10/dist-packages (from bert-score==0.3.13->axolotl==0.1) (23.1)
Collecting dill (from evaluate==0.4.0->axolotl==0.1)
  Downloading dill-0.3.6-py3-none-any.whl (110 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 110.5/110.5 kB 15.1 MB/s eta 0:00:00
Collecting xxhash (from evaluate==0.4.0->axolotl==0.1)
  Downloading xxhash-3.2.0-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (212 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 212.5/212.5 kB 22.1 MB/s eta 0:00:00
Collecting multiprocess (from evaluate==0.4.0->axolotl==0.1)
  Downloading multiprocess-0.70.14-py310-none-any.whl (134 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 134.3/134.3 kB 16.4 MB/s eta 0:00:00
Requirement already satisfied: fsspec[http]>=2021.05.0 in /usr/local/lib/python3.10/dist-packages (from evaluate==0.4.0->axolotl==0.1) (2023.4.0)
Collecting huggingface-hub>=0.7.0 (from evaluate==0.4.0->axolotl==0.1)
  Downloading huggingface_hub-0.15.1-py3-none-any.whl (236 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 236.8/236.8 kB 22.4 MB/s eta 0:00:00
Collecting responses<0.19 (from evaluate==0.4.0->axolotl==0.1)
  Downloading responses-0.18.0-py3-none-any.whl (38 kB)
Requirement already satisfied: absl-py in /usr/local/lib/python3.10/dist-packages (from rouge-score==0.1.2->axolotl==0.1) (1.4.0)
Requirement already satisfied: nltk in /usr/local/lib/python3.10/dist-packages (from rouge-score==0.1.2->axolotl==0.1) (3.8.1)
Requirement already satisfied: six>=1.14.0 in /usr/local/lib/python3.10/dist-packages (from rouge-score==0.1.2->axolotl==0.1) (1.16.0)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.10/dist-packages (from scikit-learn==1.2.2->axolotl==0.1) (1.2.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from scikit-learn==1.2.2->axolotl==0.1) (3.1.0)
Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from accelerate->axolotl==0.1) (5.9.5)
Requirement already satisfied: pyarrow>=8.0.0 in /usr/local/lib/python3.10/dist-packages (from datasets->axolotl==0.1) (9.0.0)
Collecting aiohttp (from datasets->axolotl==0.1)
  Downloading aiohttp-3.8.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.0 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.0/1.0 MB 56.1 MB/s eta 0:00:00
Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from transformers@ git+https://github.com/huggingface/transformers.git->axolotl==0.1) (3.12.0)
Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers@ git+https://github.com/huggingface/transformers.git->axolotl==0.1) (2022.10.31)
Collecting tokenizers!=0.11.3,<0.14,>=0.11.1 (from transformers@ git+https://github.com/huggingface/transformers.git->axolotl==0.1)
  Downloading tokenizers-0.13.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (7.8 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 7.8/7.8 MB 115.6 MB/s eta 0:00:00
Collecting safetensors>=0.3.1 (from transformers@ git+https://github.com/huggingface/transformers.git->axolotl==0.1)
  Downloading safetensors-0.3.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.3 MB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 1.3/1.3 MB 77.3 MB/s eta 0:00:00
Requirement already satisfied: termcolor in /usr/local/lib/python3.10/dist-packages (from fire->axolotl==0.1) (2.3.0)
Requirement already satisfied: Click!=8.0.0,>=7.0 in /usr/local/lib/python3.10/dist-packages (from wandb->axolotl==0.1) (8.1.3)
Collecting GitPython!=3.1.29,>=1.0.0 (from wandb->axolotl==0.1)
  Downloading GitPython-3.1.31-py3-none-any.whl (184 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 184.3/184.3 kB 24.5 MB/s eta 0:00:00
Collecting sentry-sdk>=1.0.0 (from wandb->axolotl==0.1)
  Downloading sentry_sdk-1.25.1-py2.py3-none-any.whl (206 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 206.7/206.7 kB 25.5 MB/s eta 0:00:00
Collecting docker-pycreds>=0.4.0 (from wandb->axolotl==0.1)
  Downloading docker_pycreds-0.4.0-py2.py3-none-any.whl (9.0 kB)
Collecting pathtools (from wandb->axolotl==0.1)
  Downloading pathtools-0.1.2.tar.gz (11 kB)
  Preparing metadata (setup.py) ... done
Collecting setproctitle (from wandb->axolotl==0.1)
  Downloading setproctitle-1.3.2-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl (30 kB)
Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from wandb->axolotl==0.1) (67.7.2)
Requirement already satisfied: appdirs>=1.4.3 in /usr/local/lib/python3.10/dist-packages (from wandb->axolotl==0.1) (1.4.4)
Requirement already satisfied: protobuf!=4.21.0,<5,>=3.19.0 in /usr/local/lib/python3.10/dist-packages (from wandb->axolotl==0.1) (3.20.3)
Collecting pyre-extensions==0.0.29 (from xformers->axolotl==0.1)
  Downloading pyre_extensions-0.0.29-py3-none-any.whl (12 kB)
Collecting typing-inspect (from pyre-extensions==0.0.29->xformers->axolotl==0.1)
  Downloading typing_inspect-0.9.0-py3-none-any.whl (8.8 kB)
Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from pyre-extensions==0.0.29->xformers->axolotl==0.1) (4.5.0)
Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (1.11.1)
Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (3.1)
Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (3.1.2)
Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.10/dist-packages (from torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (2.0.0)
Requirement already satisfied: cmake in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (3.25.2)
Requirement already satisfied: lit in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (16.0.5)
Requirement already satisfied: attrs>=17.3.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->datasets->axolotl==0.1) (23.1.0)
Requirement already satisfied: charset-normalizer<4.0,>=2.0 in /usr/local/lib/python3.10/dist-packages (from aiohttp->datasets->axolotl==0.1) (2.0.12)
Collecting multidict<7.0,>=4.5 (from aiohttp->datasets->axolotl==0.1)
  Downloading multidict-6.0.4-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (114 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 114.5/114.5 kB 15.1 MB/s eta 0:00:00
Collecting async-timeout<5.0,>=4.0.0a3 (from aiohttp->datasets->axolotl==0.1)
  Downloading async_timeout-4.0.2-py3-none-any.whl (5.8 kB)
Collecting yarl<2.0,>=1.0 (from aiohttp->datasets->axolotl==0.1)
  Downloading yarl-1.9.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (268 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 268.8/268.8 kB 32.4 MB/s eta 0:00:00
Collecting frozenlist>=1.1.1 (from aiohttp->datasets->axolotl==0.1)
  Downloading frozenlist-1.3.3-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl (149 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 149.6/149.6 kB 20.7 MB/s eta 0:00:00
Collecting aiosignal>=1.1.2 (from aiohttp->datasets->axolotl==0.1)
  Downloading aiosignal-1.3.1-py3-none-any.whl (7.6 kB)
Collecting gitdb<5,>=4.0.1 (from GitPython!=3.1.29,>=1.0.0->wandb->axolotl==0.1)
  Downloading gitdb-4.0.10-py3-none-any.whl (62 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 62.7/62.7 kB 9.6 MB/s eta 0:00:00
Requirement already satisfied: python-dateutil>=2.8.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.0.1->bert-score==0.3.13->axolotl==0.1) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.10/dist-packages (from pandas>=1.0.1->bert-score==0.3.13->axolotl==0.1) (2022.7.1)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->bert-score==0.3.13->axolotl==0.1) (1.26.15)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->bert-score==0.3.13->axolotl==0.1) (2022.12.7)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->bert-score==0.3.13->axolotl==0.1) (3.4)
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (1.0.7)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (0.11.0)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (4.39.3)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (1.4.4)
Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (8.4.0)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib->bert-score==0.3.13->axolotl==0.1) (3.0.9)
Collecting smmap<6,>=3.0.1 (from gitdb<5,>=4.0.1->GitPython!=3.1.29,>=1.0.0->wandb->axolotl==0.1)
  Downloading smmap-5.0.0-py3-none-any.whl (24 kB)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (2.1.2)
Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch>=1.0.0->bert-score==0.3.13->axolotl==0.1) (1.3.0)
Collecting mypy-extensions>=0.3.0 (from typing-inspect->pyre-extensions==0.0.29->xformers->axolotl==0.1)
  Downloading mypy_extensions-1.0.0-py3-none-any.whl (4.7 kB)
Building wheels for collected packages: rouge-score, transformers, fire, pathtools
  Building wheel for rouge-score (setup.py) ... done
  Created wheel for rouge-score: filename=rouge_score-0.1.2-py3-none-any.whl size=24934 sha256=4f4dfa3de237b24376b3e253c0284d106034f2086de524eff974fec08528e009
  Stored in directory: /root/.cache/pip/wheels/5f/dd/89/461065a73be61a532ff8599a28e9beef17985c9e9c31e541b4
  Building wheel for transformers (pyproject.toml) ... done
  Created wheel for transformers: filename=transformers-4.31.0.dev0-py3-none-any.whl size=7170473 sha256=164dfc7017620d58a037edc7544077dffe3b5500ed73f73567e8261748a2c55c
  Stored in directory: /tmp/pip-ephem-wheel-cache-9jc0rsg4/wheels/e7/9c/5b/e1a9c8007c343041e61cc484433d512ea9274272e3fcbe7c16
  Building wheel for fire (setup.py) ... done
  Created wheel for fire: filename=fire-0.5.0-py2.py3-none-any.whl size=116932 sha256=2f59363ebbfbcbc1f0f4d0d6ebb5b0d101808f66163050b92f37574e1660821a
  Stored in directory: /root/.cache/pip/wheels/90/d4/f7/9404e5db0116bd4d43e5666eaa3e70ab53723e1e3ea40c9a95
  Building wheel for pathtools (setup.py) ... done
  Created wheel for pathtools: filename=pathtools-0.1.2-py3-none-any.whl size=8791 sha256=dd6267f4f59e97d4e37762c1661146984b6d3dca2cd262a7ede8b09af42b0709
  Stored in directory: /root/.cache/pip/wheels/e7/f3/22/152153d6eb222ee7a56ff8617d80ee5207207a8c00a7aab794
Successfully built rouge-score transformers fire pathtools
Installing collected packages: tokenizers, sentencepiece, safetensors, pathtools, bitsandbytes, addict, xxhash, smmap, setproctitle, sentry-sdk, mypy-extensions, multidict, frozenlist, fire, einops, docker-pycreds, dill, async-timeout, yarl, typing-inspect, rouge-score, responses, multiprocess, huggingface-hub, gitdb, aiosignal, transformers, pyre-extensions, GitPython, aiohttp, wandb, datasets, evaluate, xformers, bert-score, accelerate, axolotl
  Running setup.py develop for axolotl
Successfully installed GitPython-3.1.31 accelerate-0.20.3 addict-2.4.0 aiohttp-3.8.4 aiosignal-1.3.1 async-timeout-4.0.2 axolotl-0.1 bert-score-0.3.13 bitsandbytes-0.39.0 datasets-2.12.0 dill-0.3.6 docker-pycreds-0.4.0 einops-0.6.1 evaluate-0.4.0 fire-0.5.0 frozenlist-1.3.3 gitdb-4.0.10 huggingface-hub-0.15.1 multidict-6.0.4 multiprocess-0.70.14 mypy-extensions-1.0.0 pathtools-0.1.2 pyre-extensions-0.0.29 responses-0.18.0 rouge-score-0.1.2 safetensors-0.3.1 sentencepiece-0.1.99 sentry-sdk-1.25.1 setproctitle-1.3.2 smmap-5.0.0 tokenizers-0.13.3 transformers-4.31.0.dev0 typing-inspect-0.9.0 wandb-0.15.4 xformers-0.0.20 xxhash-3.2.0 yarl-1.9.2

!pip install -U git+https://github.com/huggingface/peft.git

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting git+https://github.com/huggingface/peft.git
  Cloning https://github.com/huggingface/peft.git to /tmp/pip-req-build-vajurb24
  Running command git clone --filter=blob:none --quiet https://github.com/huggingface/peft.git /tmp/pip-req-build-vajurb24
  Resolved https://github.com/huggingface/peft.git to commit 189a6b8e357ecda05ccde13999e4c35759596a67
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... done
Requirement already satisfied: numpy>=1.17 in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (1.22.4)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (23.1)
Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (5.9.5)
Requirement already satisfied: pyyaml in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (6.0)
Requirement already satisfied: torch>=1.13.0 in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (2.0.1+cu118)
Requirement already satisfied: transformers in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (4.31.0.dev0)
Requirement already satisfied: accelerate in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (0.20.3)
Requirement already satisfied: safetensors in /usr/local/lib/python3.10/dist-packages (from peft==0.4.0.dev0) (0.3.1)
Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (3.12.0)
Requirement already satisfied: typing-extensions in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (4.5.0)
Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (1.11.1)
Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (3.1)
Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (3.1.2)
Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.10/dist-packages (from torch>=1.13.0->peft==0.4.0.dev0) (2.0.0)
Requirement already satisfied: cmake in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=1.13.0->peft==0.4.0.dev0) (3.25.2)
Requirement already satisfied: lit in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch>=1.13.0->peft==0.4.0.dev0) (16.0.5)
Requirement already satisfied: huggingface-hub<1.0,>=0.14.1 in /usr/local/lib/python3.10/dist-packages (from transformers->peft==0.4.0.dev0) (0.15.1)
Requirement already satisfied: regex!=2019.12.17 in /usr/local/lib/python3.10/dist-packages (from transformers->peft==0.4.0.dev0) (2022.10.31)
Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from transformers->peft==0.4.0.dev0) (2.27.1)
Requirement already satisfied: tokenizers!=0.11.3,<0.14,>=0.11.1 in /usr/local/lib/python3.10/dist-packages (from transformers->peft==0.4.0.dev0) (0.13.3)
Requirement already satisfied: tqdm>=4.27 in /usr/local/lib/python3.10/dist-packages (from transformers->peft==0.4.0.dev0) (4.65.0)
Requirement already satisfied: fsspec in /usr/local/lib/python3.10/dist-packages (from huggingface-hub<1.0,>=0.14.1->transformers->peft==0.4.0.dev0) (2023.4.0)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch>=1.13.0->peft==0.4.0.dev0) (2.1.2)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->peft==0.4.0.dev0) (1.26.15)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->peft==0.4.0.dev0) (2022.12.7)
Requirement already satisfied: charset-normalizer~=2.0.0 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->peft==0.4.0.dev0) (2.0.12)
Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->transformers->peft==0.4.0.dev0) (3.4)
Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch>=1.13.0->peft==0.4.0.dev0) (1.3.0)
Building wheels for collected packages: peft
  Building wheel for peft (pyproject.toml) ... done
  Created wheel for peft: filename=peft-0.4.0.dev0-py3-none-any.whl size=59308 sha256=414b6cadbcf4223f6eb472b8145ff607431f62b70e86d9b45b7ecced17ddd36c
  Stored in directory: /tmp/pip-ephem-wheel-cache-5vh9a2_r/wheels/d7/c7/de/1368fac8590e1b103ddc2ec2a28ad51d83aded1a3830e8a087
Successfully built peft
Installing collected packages: peft
Successfully installed peft-0.4.0.dev0

!accelerate config --config_file configs/accelerate/default_config.yaml default

2023-06-10 11:16:08.926713: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
accelerate configuration saved at /root/.cache/huggingface/accelerate/default_config.yaml

%%writefile examples/falcon/config-1b-qlora.yml
# 1b: tiiuae/falcon-rw-1b
# 40b: tiiuae/falcon-40b
base_model: tiiuae/falcon-rw-1b
base_model_config: tiiuae/falcon-rw-1b
# required by falcon custom model code: https://huggingface.co/tiiuae/falcon-rw-1b/tree/main
trust_remote_code: true
model_type: AutoModelForCausalLM
tokenizer_type: AutoTokenizer
load_in_8bit: false
# enable 4bit for QLoRA
load_in_4bit: true
gptq: false
strict: false
push_dataset_to_hub:
datasets:
  - path: QingyiSi/Alpaca-CoT
    data_files:
      - Chain-of-Thought/formatted_cot_data/gsm8k_train.json
    type: "alpaca:chat"
dataset_prepared_path: last_run_prepared
val_set_size: 0.01
# enable QLoRA
adapter: qlora
lora_model_dir:
sequence_len: 2048
max_packed_sequence_len:

# hyperparameters from QLoRA paper Appendix B.2
# "We find hyperparameters to be largely robust across datasets"
lora_r: 64
lora_alpha: 16
# 0.1 for models up to 13B
# 0.05 for 33B and 65B models
lora_dropout: 0.05
# add LoRA modules on all linear layers of the base model
lora_target_modules:
lora_target_linear: true
lora_fan_in_fan_out:

wandb_project:
wandb_watch:
wandb_run_id:
wandb_log_model:
output_dir: ./qlora-out

some_config: some_string

# QLoRA paper Table 9
# - 16 for 7b & 13b
# - 32 for 33b, 64 for 64b
# Max size tested on A6000
# - 7b: 40
# - 40b: 4
# decrease if OOM, increase for max VRAM utilization
micro_batch_size: 1
gradient_accumulation_steps: 1
num_epochs: 3
# Optimizer for QLoRA
optimizer: paged_adamw_32bit
torchdistx_path:
lr_scheduler: cosine
# QLoRA paper Table 9
# - 2e-4 for 7b & 13b
# - 1e-4 for 33b & 64b
learning_rate: 0.0002
train_on_inputs: false
group_by_length: false
bf16: false
fp16: false
tf32: false
gradient_checkpointing: true
# stop training after this many evaluation losses have increased in a row
# https://huggingface.co/transformers/v4.2.2/_modules/transformers/trainer_callback.html#EarlyStoppingCallback
early_stopping_patience: 3
resume_from_checkpoint:
auto_resume_from_checkpoints: true
local_rank:
logging_steps: 1
xformers_attention: true
flash_attention:
gptq_groupsize:
gptq_model_v1:
warmup_steps: 10
eval_steps: 5
save_steps: 10
debug:
deepspeed:
weight_decay: 0.000001
fsdp:
fsdp_config:
special_tokens:
  pad_token: "<|endoftext|>"
  bos_token: ">>ABSTRACT<<"
  eos_token: "<|endoftext|>"

Overwriting examples/falcon/config-1b-qlora.yml

%%writefile ds_config.json
{
  "zero_optimization": {
    "stage": 3,
    "offload_optimizer": {
      "device": "cpu",
      "pin_memory": true
    },
    "offload_param": {
      "device": "cpu",
      "pin_memory": true
    },
    "overlap_comm": true,
    "contiguous_gradients": true,
    "sub_group_size": 0,
    "reduce_bucket_size": "auto",
    "stage3_prefetch_bucket_size": "auto",
    "stage3_param_persistence_threshold": "auto",
    "stage3_max_live_parameters": 0,
    "stage3_max_reuse_distance": 0,
    "stage3_gather_16bit_weights_on_model_save": true
  },
  "bf16": {
    "enabled": "auto"
  },
  "fp16": {
    "enabled": "auto",
    "auto_cast": false,
    "loss_scale": 0,
    "initial_scale_power": 32,
    "loss_scale_window": 1000,
    "hysteresis": 2,
    "min_loss_scale": 1
  },
  "optimizer": {
    "type": "AdamW",
    "params": {
      "lr": "auto",
      "betas": [
        0.9,
        0.999
      ],
      "eps": 1e-8,
      "weight_decay": "auto"
    }
  },
  "gradient_accumulation_steps": "auto",
  "train_micro_batch_size_per_gpu": "auto",
  "wall_clock_breakdown": false
}

Overwriting ds_config.json

%%writefile scripts/ft.py
import os
from pathlib import Path
import fire
import logging
import finetune
from axolotl.utils.trainer import setup_trainer as setup_trainer_orig

logging.basicConfig(level=os.getenv("LOG_LEVEL", "INFO"))

def train_ex(
    config: Path = Path("configs/"),
    prepare_ds_only: bool = False,
    **kwargs,
):
  logging.info('train_ex before')
  finetune.train(config, prepare_ds_only, **kwargs)
  logging.info('train_ex after')

def setup_trainer_ex(cfg, train_dataset, eval_dataset, model, tokenizer):
  logging.info('setup_trainer_ex before')
  logging.info(f'cfg.some_config = {cfg.some_config}')
  trainer = setup_trainer_orig(cfg, train_dataset, eval_dataset, model, tokenizer)
  logging.info('setup_trainer_ex after')
  return trainer

finetune.setup_trainer = setup_trainer_ex

if __name__ == "__main__":
    fire.Fire(train_ex)

Overwriting scripts/ft.py

%env ACCELERATE_USE_DEEPSPEED=true

env: ACCELERATE_USE_DEEPSPEED=true

!pip install deepspeed

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting deepspeed
  Downloading deepspeed-0.9.4.tar.gz (808 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 808.8/808.8 kB 14.7 MB/s eta 0:00:00
  Preparing metadata (setup.py) ... done
Collecting hjson (from deepspeed)
  Downloading hjson-3.1.0-py3-none-any.whl (54 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 54.0/54.0 kB 7.1 MB/s eta 0:00:00
Collecting ninja (from deepspeed)
  Downloading ninja-1.11.1-py2.py3-none-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (145 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 146.0/146.0 kB 18.7 MB/s eta 0:00:00
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from deepspeed) (1.22.4)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from deepspeed) (23.1)
Requirement already satisfied: psutil in /usr/local/lib/python3.10/dist-packages (from deepspeed) (5.9.5)
Requirement already satisfied: py-cpuinfo in /usr/local/lib/python3.10/dist-packages (from deepspeed) (9.0.0)
Requirement already satisfied: pydantic<2.0.0 in /usr/local/lib/python3.10/dist-packages (from deepspeed) (1.10.7)
Requirement already satisfied: torch in /usr/local/lib/python3.10/dist-packages (from deepspeed) (2.0.1+cu118)
Requirement already satisfied: tqdm in /usr/local/lib/python3.10/dist-packages (from deepspeed) (4.65.0)
Requirement already satisfied: typing-extensions>=4.2.0 in /usr/local/lib/python3.10/dist-packages (from pydantic<2.0.0->deepspeed) (4.5.0)
Requirement already satisfied: filelock in /usr/local/lib/python3.10/dist-packages (from torch->deepspeed) (3.12.0)
Requirement already satisfied: sympy in /usr/local/lib/python3.10/dist-packages (from torch->deepspeed) (1.11.1)
Requirement already satisfied: networkx in /usr/local/lib/python3.10/dist-packages (from torch->deepspeed) (3.1)
Requirement already satisfied: jinja2 in /usr/local/lib/python3.10/dist-packages (from torch->deepspeed) (3.1.2)
Requirement already satisfied: triton==2.0.0 in /usr/local/lib/python3.10/dist-packages (from torch->deepspeed) (2.0.0)
Requirement already satisfied: cmake in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch->deepspeed) (3.25.2)
Requirement already satisfied: lit in /usr/local/lib/python3.10/dist-packages (from triton==2.0.0->torch->deepspeed) (16.0.5)
Requirement already satisfied: MarkupSafe>=2.0 in /usr/local/lib/python3.10/dist-packages (from jinja2->torch->deepspeed) (2.1.2)
Requirement already satisfied: mpmath>=0.19 in /usr/local/lib/python3.10/dist-packages (from sympy->torch->deepspeed) (1.3.0)
Building wheels for collected packages: deepspeed
  Building wheel for deepspeed (setup.py) ... done
  Created wheel for deepspeed: filename=deepspeed-0.9.4-py3-none-any.whl size=843997 sha256=98941ac1570bb2277f4920544479cbb77a8b7fc28f3f11f2a350115652c4e833
  Stored in directory: /root/.cache/pip/wheels/2d/ae/38/1d1c49ac8687c5808b3732e3541b6c896459fb8404763eb98b
Successfully built deepspeed
Installing collected packages: ninja, hjson, deepspeed
Successfully installed deepspeed-0.9.4 hjson-3.1.0 ninja-1.11.1

!rm /root/.cache/huggingface/accelerate/default_config.yaml

!accelerate launch scripts/finetune.py examples/falcon/config-1b-qlora.yml  --deepspeed ds_config.json

[2023-06-10 15:08:08,918] [INFO] [real_accelerator.py:110:get_accelerator] Setting ds_accelerator to cuda (auto detect)
2023-06-10 15:08:11.874584: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
[15:08:13] WARNING  The following values were not passed to        ]8;id=807900;file:///usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py\launch.py]8;;\:]8;id=572740;file:///usr/local/lib/python3.10/dist-packages/accelerate/commands/launch.py#913\913]8;;\
                    `accelerate launch` and had defaults used                   
                    instead:                                                    
                            `--num_processes` was set to a value                
                    of `1`                                                      
                            `--num_machines` was set to a value of              
                    `1`                                                         
                            `--mixed_precision` was set to a value              
                    of `'no'`                                                   
                            `--dynamo_backend` was set to a value               
                    of `'no'`                                                   
                    To avoid this warning pass in values for each               
                    of the problematic parameters or run                        
                    `accelerate config`.                                        

===================================BUG REPORT===================================
Welcome to bitsandbytes. For bug reports, please run

python -m bitsandbytes

 and submit this information together with your error trace to: https://github.com/TimDettmers/bitsandbytes/issues
================================================================================
bin /usr/local/lib/python3.10/dist-packages/bitsandbytes/libbitsandbytes_cuda118.so
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: /usr/lib64-nvidia did not contain ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] as expected! Searching further paths...
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('/sys/fs/cgroup/memory.events /var/colab/cgroup/jupyter-children/memory.events')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('8013'), PosixPath('http'), PosixPath('//172.28.0.1')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('//colab.research.google.com/tun/m/cc48301118ce562b961b3c22d803539adc1e0c19/gpu-t4-s-tccs3c5dv19r --tunnel_background_save_delay=10s --tunnel_periodic_background_save_frequency=30m0s --enable_output_coalescing=true --output_coalescing_required=true'), PosixPath('--logtostderr --listen_host=172.28.0.12 --target_host=172.28.0.12 --tunnel_background_save_url=https')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('/env/python')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('//ipykernel.pylab.backend_inline'), PosixPath('module')}
  warn(msg)
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching in backup paths...
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: Found duplicate ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] files: {PosixPath('/usr/local/cuda/lib64/libcudart.so.11.0'), PosixPath('/usr/local/cuda/lib64/libcudart.so')}.. We'll flip a coin and try one of these, in order to fail forward.
Either way, this might cause trouble in the future:
If you get `CUDA error: invalid device function` errors, the above might be the cause and the solution is to make sure only one ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] in the paths that we search based on your env.
  warn(msg)
CUDA SETUP: CUDA runtime path found: /usr/local/cuda/lib64/libcudart.so.11.0
CUDA SETUP: Highest compute capability among GPUs detected: 7.5
CUDA SETUP: Detected CUDA version 118
CUDA SETUP: Loading binary /usr/local/lib/python3.10/dist-packages/bitsandbytes/libbitsandbytes_cuda118.so...
[2023-06-10 15:08:17,560] [INFO] [real_accelerator.py:110:get_accelerator] Setting ds_accelerator to cuda (auto detect)
2023-06-10 15:08:20.031162: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
WARNING:root:`trust_remote_code` is set to true. Please make sure that you reviewed the remote code/model.
INFO:root:loading tokenizer... tiiuae/falcon-rw-1b
Using pad_token, but it is not set yet.
INFO:root:Loading prepared dataset from disk at last_run_prepared/0ecc5b78e3ce4254b22e749b093712b4...
INFO:root:Prepared dataset loaded from disk...
INFO:root:loading model and peft_config...
INFO:root:converting PEFT model w/ prepare_model_for_kbit_training
INFO:root:found linear modules: ['dense_h_to_4h', 'query_key_value', 'dense', 'dense_4h_to_h']
trainable params: 50,331,648 || all params: 757,911,552 || trainable%: 6.6408339953630895
Using the `WANDB_DISABLED` environment variable is deprecated and will be removed in v5. Use the --report_to flag to control the integrations used for logging result (for instance --report_to none).
INFO:root:Compiling torch model
INFO:root:Pre-saving adapter config to ./qlora-out
INFO:root:Starting trainer...
  0% 0/22194 [00:00<?, ?it/s]You're using a GPT2TokenizerFast tokenizer. Please note that with a fast tokenizer, using the `__call__` method is faster than using a method to encode the text followed by a call to the `pad` method to get a padded encoding.
{'loss': 1.6302, 'learning_rate': 2e-05, 'epoch': 0.0}
{'loss': 2.0857, 'learning_rate': 4e-05, 'epoch': 0.0}
{'loss': 1.9553, 'learning_rate': 6e-05, 'epoch': 0.0}
{'loss': 1.3392, 'learning_rate': 8e-05, 'epoch': 0.0}
{'loss': 1.4295, 'learning_rate': 0.0001, 'epoch': 0.0}
  0% 5/22194 [00:04<5:34:00,  1.11it/s]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.76it/s]
  5% 4/75 [00:00<00:11,  6.16it/s]
  7% 5/75 [00:00<00:13,  5.06it/s]
  8% 6/75 [00:01<00:14,  4.86it/s]
  9% 7/75 [00:01<00:15,  4.27it/s]
 11% 8/75 [00:01<00:16,  4.16it/s]
 12% 9/75 [00:01<00:15,  4.28it/s]
 13% 10/75 [00:02<00:15,  4.24it/s]
 15% 11/75 [00:02<00:14,  4.43it/s]
 16% 12/75 [00:02<00:14,  4.34it/s]
 17% 13/75 [00:02<00:14,  4.17it/s]
 19% 14/75 [00:03<00:14,  4.16it/s]
 20% 15/75 [00:03<00:14,  4.27it/s]
 21% 16/75 [00:03<00:14,  4.03it/s]
 23% 17/75 [00:03<00:14,  4.05it/s]
 24% 18/75 [00:04<00:13,  4.23it/s]
 25% 19/75 [00:04<00:13,  4.08it/s]
 27% 20/75 [00:04<00:13,  4.00it/s]
 28% 21/75 [00:04<00:13,  3.94it/s]
 29% 22/75 [00:05<00:13,  3.83it/s]
 31% 23/75 [00:05<00:14,  3.69it/s]
 32% 24/75 [00:05<00:13,  3.81it/s]
 33% 25/75 [00:05<00:12,  4.16it/s]
 35% 26/75 [00:06<00:11,  4.31it/s]
 36% 27/75 [00:06<00:11,  4.16it/s]
 37% 28/75 [00:06<00:11,  4.05it/s]
 39% 29/75 [00:06<00:11,  3.91it/s]
 40% 30/75 [00:07<00:12,  3.68it/s]
 41% 31/75 [00:07<00:11,  3.68it/s]
 43% 32/75 [00:07<00:10,  3.95it/s]
 44% 33/75 [00:07<00:10,  3.89it/s]
 45% 34/75 [00:08<00:10,  3.73it/s]
 47% 35/75 [00:08<00:11,  3.63it/s]
 48% 36/75 [00:08<00:10,  3.65it/s]
 49% 37/75 [00:09<00:10,  3.66it/s]
 51% 38/75 [00:09<00:10,  3.69it/s]
 52% 39/75 [00:09<00:09,  3.69it/s]
 53% 40/75 [00:09<00:09,  3.74it/s]
 55% 41/75 [00:10<00:09,  3.73it/s]
 56% 42/75 [00:10<00:09,  3.66it/s]
 57% 43/75 [00:10<00:08,  3.71it/s]
 59% 44/75 [00:10<00:07,  4.07it/s]
 60% 45/75 [00:11<00:07,  3.87it/s]
 61% 46/75 [00:11<00:08,  3.59it/s]
 63% 47/75 [00:11<00:08,  3.41it/s]
 64% 48/75 [00:12<00:08,  3.30it/s]
 65% 49/75 [00:12<00:07,  3.46it/s]
 67% 50/75 [00:12<00:07,  3.41it/s]
 68% 51/75 [00:12<00:06,  3.49it/s]
 69% 52/75 [00:13<00:06,  3.68it/s]
 71% 53/75 [00:13<00:06,  3.23it/s]
 72% 54/75 [00:13<00:06,  3.17it/s]
 73% 55/75 [00:14<00:05,  3.55it/s]
 75% 56/75 [00:14<00:05,  3.53it/s]
 76% 57/75 [00:14<00:05,  3.38it/s]
 77% 58/75 [00:14<00:04,  3.53it/s]
 79% 59/75 [00:15<00:04,  3.84it/s]
 80% 60/75 [00:15<00:04,  3.61it/s]
 81% 61/75 [00:15<00:03,  3.53it/s]
 83% 62/75 [00:16<00:03,  3.38it/s]
 84% 63/75 [00:16<00:03,  3.43it/s]
 85% 64/75 [00:16<00:03,  3.48it/s]
 87% 65/75 [00:16<00:02,  3.59it/s]
 88% 66/75 [00:17<00:02,  3.97it/s]
 89% 67/75 [00:17<00:01,  4.05it/s]
 91% 68/75 [00:17<00:01,  3.84it/s]
 92% 69/75 [00:17<00:01,  3.85it/s]
 93% 70/75 [00:18<00:01,  4.06it/s]
 95% 71/75 [00:18<00:01,  3.92it/s]
 96% 72/75 [00:18<00:00,  3.89it/s]
 97% 73/75 [00:18<00:00,  3.92it/s]
 99% 74/75 [00:19<00:00,  3.69it/s]
                                       
{'eval_loss': 1.944600224494934, 'eval_runtime': 19.7817, 'eval_samples_per_second': 3.791, 'eval_steps_per_second': 3.791, 'epoch': 0.0}
  0% 5/22194 [00:24<5:34:00,  1.11it/s]
100% 75/75 [00:19<00:00,  3.62it/s]
{'loss': 1.8146, 'learning_rate': 0.00012, 'epoch': 0.0}
{'loss': 1.5587, 'learning_rate': 0.00014, 'epoch': 0.0}
{'loss': 1.914, 'learning_rate': 0.00016, 'epoch': 0.0}
{'loss': 1.4058, 'learning_rate': 0.00018, 'epoch': 0.0}
{'loss': 1.9108, 'learning_rate': 0.0002, 'epoch': 0.0}
  0% 10/22194 [00:29<14:26:00,  2.34s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:08,  9.06it/s]
  4% 3/75 [00:00<00:11,  6.19it/s]
  5% 4/75 [00:00<00:14,  4.75it/s]
  7% 5/75 [00:01<00:16,  4.13it/s]
  8% 6/75 [00:01<00:17,  4.01it/s]
  9% 7/75 [00:01<00:19,  3.57it/s]
 11% 8/75 [00:01<00:19,  3.49it/s]
 12% 9/75 [00:02<00:18,  3.66it/s]
 13% 10/75 [00:02<00:17,  3.62it/s]
 15% 11/75 [00:02<00:16,  3.92it/s]
 16% 12/75 [00:02<00:16,  3.79it/s]
 17% 13/75 [00:03<00:16,  3.66it/s]
 19% 14/75 [00:03<00:16,  3.64it/s]
 20% 15/75 [00:03<00:15,  3.81it/s]
 21% 16/75 [00:04<00:16,  3.57it/s]
 23% 17/75 [00:04<00:16,  3.58it/s]
 24% 18/75 [00:04<00:14,  3.81it/s]
 25% 19/75 [00:04<00:15,  3.63it/s]
 27% 20/75 [00:05<00:15,  3.51it/s]
 28% 21/75 [00:05<00:15,  3.46it/s]
 29% 22/75 [00:05<00:15,  3.45it/s]
 31% 23/75 [00:06<00:15,  3.29it/s]
 32% 24/75 [00:06<00:15,  3.33it/s]
 33% 25/75 [00:06<00:13,  3.62it/s]
 35% 26/75 [00:06<00:12,  3.87it/s]
 36% 27/75 [00:07<00:12,  3.70it/s]
 37% 28/75 [00:07<00:13,  3.56it/s]
 39% 29/75 [00:07<00:13,  3.48it/s]
 40% 30/75 [00:08<00:13,  3.29it/s]
 41% 31/75 [00:08<00:13,  3.29it/s]
 43% 32/75 [00:08<00:12,  3.50it/s]
 44% 33/75 [00:08<00:12,  3.47it/s]
 45% 34/75 [00:09<00:12,  3.35it/s]
 47% 35/75 [00:09<00:12,  3.28it/s]
 48% 36/75 [00:09<00:11,  3.29it/s]
 49% 37/75 [00:10<00:11,  3.30it/s]
 51% 38/75 [00:10<00:11,  3.34it/s]
 52% 39/75 [00:10<00:10,  3.35it/s]
 53% 40/75 [00:11<00:10,  3.37it/s]
 55% 41/75 [00:11<00:10,  3.39it/s]
 56% 42/75 [00:11<00:09,  3.37it/s]
 57% 43/75 [00:11<00:09,  3.48it/s]
 59% 44/75 [00:12<00:08,  3.65it/s]
 60% 45/75 [00:12<00:08,  3.50it/s]
 61% 46/75 [00:12<00:08,  3.27it/s]
 63% 47/75 [00:13<00:09,  3.11it/s]
 64% 48/75 [00:13<00:08,  3.11it/s]
 65% 49/75 [00:13<00:08,  3.15it/s]
 67% 50/75 [00:14<00:07,  3.13it/s]
 68% 51/75 [00:14<00:07,  3.27it/s]
 69% 52/75 [00:14<00:06,  3.38it/s]
 71% 53/75 [00:15<00:07,  3.00it/s]
 72% 54/75 [00:15<00:06,  3.05it/s]
 73% 55/75 [00:15<00:06,  3.32it/s]
 75% 56/75 [00:15<00:05,  3.29it/s]
 76% 57/75 [00:16<00:05,  3.25it/s]
 77% 58/75 [00:16<00:04,  3.43it/s]
 79% 59/75 [00:16<00:04,  3.56it/s]
 80% 60/75 [00:17<00:04,  3.41it/s]
 81% 61/75 [00:17<00:04,  3.35it/s]
 83% 62/75 [00:17<00:03,  3.26it/s]
 84% 63/75 [00:18<00:03,  3.28it/s]
 85% 64/75 [00:18<00:03,  3.30it/s]
 87% 65/75 [00:18<00:02,  3.40it/s]
 88% 66/75 [00:18<00:02,  3.78it/s]
 89% 67/75 [00:19<00:02,  3.97it/s]
 91% 68/75 [00:19<00:01,  3.76it/s]
 92% 69/75 [00:19<00:01,  3.68it/s]
 93% 70/75 [00:19<00:01,  3.98it/s]
 95% 71/75 [00:20<00:01,  3.87it/s]
 96% 72/75 [00:20<00:00,  3.77it/s]
 97% 73/75 [00:20<00:00,  3.91it/s]
 99% 74/75 [00:20<00:00,  3.65it/s]
                                         
{'eval_loss': 1.7404545545578003, 'eval_runtime': 21.5061, 'eval_samples_per_second': 3.487, 'eval_steps_per_second': 3.487, 'epoch': 0.0}
  0% 10/22194 [00:50<14:26:00,  2.34s/it]
100% 75/75 [00:21<00:00,  3.59it/s]
{'loss': 2.2223, 'learning_rate': 0.0001999999989972561, 'epoch': 0.0}
{'loss': 1.9906, 'learning_rate': 0.00019999999598902438, 'epoch': 0.0}
{'loss': 1.9828, 'learning_rate': 0.00019999999097530487, 'epoch': 0.0}
{'loss': 1.358, 'learning_rate': 0.00019999998395609775, 'epoch': 0.0}
{'loss': 1.1241, 'learning_rate': 0.0001999999749314031, 'epoch': 0.0}
  0% 15/22194 [00:57<17:29:46,  2.84s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.82it/s]
  5% 4/75 [00:00<00:13,  5.24it/s]
  7% 5/75 [00:01<00:15,  4.48it/s]
  8% 6/75 [00:01<00:15,  4.45it/s]
  9% 7/75 [00:01<00:17,  3.91it/s]
 11% 8/75 [00:01<00:18,  3.71it/s]
 12% 9/75 [00:02<00:16,  3.94it/s]
 13% 10/75 [00:02<00:16,  3.99it/s]
 15% 11/75 [00:02<00:15,  4.13it/s]
 16% 12/75 [00:02<00:15,  3.99it/s]
 17% 13/75 [00:03<00:15,  3.90it/s]
 19% 14/75 [00:03<00:15,  3.99it/s]
 20% 15/75 [00:03<00:14,  4.03it/s]
 21% 16/75 [00:03<00:15,  3.75it/s]
 23% 17/75 [00:04<00:15,  3.84it/s]
 24% 18/75 [00:04<00:14,  4.05it/s]
 25% 19/75 [00:04<00:14,  3.87it/s]
 27% 20/75 [00:04<00:14,  3.73it/s]
 28% 21/75 [00:05<00:14,  3.73it/s]
 29% 22/75 [00:05<00:14,  3.68it/s]
 31% 23/75 [00:05<00:14,  3.51it/s]
 32% 24/75 [00:06<00:14,  3.63it/s]
 33% 25/75 [00:06<00:12,  3.97it/s]
 35% 26/75 [00:06<00:11,  4.11it/s]
 36% 27/75 [00:06<00:12,  3.92it/s]
 37% 28/75 [00:07<00:12,  3.81it/s]
 39% 29/75 [00:07<00:12,  3.76it/s]
 40% 30/75 [00:07<00:12,  3.55it/s]
 41% 31/75 [00:07<00:12,  3.47it/s]
 43% 32/75 [00:08<00:11,  3.81it/s]
 44% 33/75 [00:08<00:11,  3.80it/s]
 45% 34/75 [00:08<00:11,  3.67it/s]
 47% 35/75 [00:08<00:11,  3.52it/s]
 48% 36/75 [00:09<00:10,  3.55it/s]
 49% 37/75 [00:09<00:11,  3.42it/s]
 51% 38/75 [00:09<00:10,  3.66it/s]
 52% 39/75 [00:10<00:09,  3.63it/s]
 53% 40/75 [00:10<00:09,  3.68it/s]
 55% 41/75 [00:10<00:09,  3.68it/s]
 56% 42/75 [00:10<00:09,  3.60it/s]
 57% 43/75 [00:11<00:08,  3.67it/s]
 59% 44/75 [00:11<00:07,  3.96it/s]
 60% 45/75 [00:11<00:07,  3.90it/s]
 61% 46/75 [00:11<00:08,  3.59it/s]
 63% 47/75 [00:12<00:08,  3.41it/s]
 64% 48/75 [00:12<00:07,  3.41it/s]
 65% 49/75 [00:12<00:07,  3.41it/s]
 67% 50/75 [00:13<00:07,  3.39it/s]
 68% 51/75 [00:13<00:06,  3.56it/s]
 69% 52/75 [00:13<00:06,  3.66it/s]
 71% 53/75 [00:14<00:06,  3.23it/s]
 72% 54/75 [00:14<00:06,  3.29it/s]
 73% 55/75 [00:14<00:05,  3.59it/s]
 75% 56/75 [00:14<00:05,  3.54it/s]
 76% 57/75 [00:15<00:05,  3.51it/s]
 77% 58/75 [00:15<00:04,  3.65it/s]
 79% 59/75 [00:15<00:04,  3.84it/s]
 80% 60/75 [00:15<00:04,  3.67it/s]
 81% 61/75 [00:16<00:03,  3.58it/s]
 83% 62/75 [00:16<00:03,  3.46it/s]
 84% 63/75 [00:16<00:03,  3.51it/s]
 85% 64/75 [00:17<00:03,  3.57it/s]
 87% 65/75 [00:17<00:02,  3.71it/s]
 88% 66/75 [00:17<00:02,  4.05it/s]
 89% 67/75 [00:17<00:01,  4.14it/s]
 91% 68/75 [00:18<00:01,  3.95it/s]
 92% 69/75 [00:18<00:01,  3.98it/s]
 93% 70/75 [00:18<00:01,  4.19it/s]
 95% 71/75 [00:18<00:00,  4.06it/s]
 96% 72/75 [00:19<00:00,  3.96it/s]
 97% 73/75 [00:19<00:00,  4.06it/s]
 99% 74/75 [00:19<00:00,  3.84it/s]
                                         
{'eval_loss': 1.6510955095291138, 'eval_runtime': 20.0901, 'eval_samples_per_second': 3.733, 'eval_steps_per_second': 3.733, 'epoch': 0.0}
  0% 15/22194 [01:17<17:29:46,  2.84s/it]
100% 75/75 [00:20<00:00,  3.78it/s]
{'loss': 1.4875, 'learning_rate': 0.00019999996390122112, 'epoch': 0.0}
{'loss': 1.7433, 'learning_rate': 0.00019999995086555203, 'epoch': 0.0}
{'loss': 1.3966, 'learning_rate': 0.0001999999358243961, 'epoch': 0.0}
{'loss': 1.073, 'learning_rate': 0.00019999991877775363, 'epoch': 0.0}
{'loss': 1.4507, 'learning_rate': 0.00019999989972562494, 'epoch': 0.0}
  0% 20/22194 [01:21<16:12:02,  2.63s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.19it/s]
  4% 3/75 [00:00<00:10,  7.00it/s]
  5% 4/75 [00:00<00:13,  5.23it/s]
  7% 5/75 [00:01<00:16,  4.13it/s]
  8% 6/75 [00:01<00:15,  4.39it/s]
  9% 7/75 [00:01<00:17,  3.88it/s]
 11% 8/75 [00:01<00:18,  3.66it/s]
 12% 9/75 [00:02<00:16,  3.95it/s]
 13% 10/75 [00:02<00:16,  3.87it/s]
 15% 11/75 [00:02<00:15,  4.19it/s]
 16% 12/75 [00:02<00:15,  4.10it/s]
 17% 13/75 [00:03<00:15,  3.90it/s]
 19% 14/75 [00:03<00:15,  3.88it/s]
 20% 15/75 [00:03<00:14,  4.18it/s]
 21% 16/75 [00:03<00:15,  3.81it/s]
 23% 17/75 [00:04<00:14,  3.92it/s]
 24% 18/75 [00:04<00:13,  4.11it/s]
 25% 19/75 [00:04<00:14,  3.95it/s]
 27% 20/75 [00:04<00:14,  3.84it/s]
 28% 21/75 [00:05<00:14,  3.79it/s]
 29% 22/75 [00:05<00:14,  3.72it/s]
 31% 23/75 [00:05<00:14,  3.57it/s]
 32% 24/75 [00:05<00:13,  3.69it/s]
 33% 25/75 [00:06<00:12,  3.97it/s]
 35% 26/75 [00:06<00:11,  4.18it/s]
 36% 27/75 [00:06<00:11,  4.06it/s]
 37% 28/75 [00:06<00:12,  3.89it/s]
 39% 29/75 [00:07<00:12,  3.79it/s]
 40% 30/75 [00:07<00:12,  3.60it/s]
 41% 31/75 [00:07<00:12,  3.59it/s]
 43% 32/75 [00:08<00:11,  3.82it/s]
 44% 33/75 [00:08<00:11,  3.80it/s]
 45% 34/75 [00:08<00:11,  3.66it/s]
 47% 35/75 [00:08<00:11,  3.54it/s]
 48% 36/75 [00:09<00:11,  3.54it/s]
 49% 37/75 [00:09<00:10,  3.58it/s]
 51% 38/75 [00:09<00:10,  3.61it/s]
 52% 39/75 [00:09<00:10,  3.58it/s]
 53% 40/75 [00:10<00:09,  3.64it/s]
 55% 41/75 [00:10<00:09,  3.66it/s]
 56% 42/75 [00:10<00:09,  3.64it/s]
 57% 43/75 [00:11<00:08,  3.76it/s]
 59% 44/75 [00:11<00:07,  3.93it/s]
 60% 45/75 [00:11<00:07,  3.79it/s]
 61% 46/75 [00:11<00:08,  3.52it/s]
 63% 47/75 [00:12<00:08,  3.35it/s]
 64% 48/75 [00:12<00:08,  3.36it/s]
 65% 49/75 [00:12<00:07,  3.39it/s]
 67% 50/75 [00:13<00:07,  3.35it/s]
 68% 51/75 [00:13<00:06,  3.51it/s]
 69% 52/75 [00:13<00:06,  3.62it/s]
 71% 53/75 [00:14<00:06,  3.22it/s]
 72% 54/75 [00:14<00:06,  3.17it/s]
 73% 55/75 [00:14<00:05,  3.60it/s]
 75% 56/75 [00:14<00:05,  3.51it/s]
 76% 57/75 [00:15<00:05,  3.41it/s]
 77% 58/75 [00:15<00:04,  3.53it/s]
 79% 59/75 [00:15<00:04,  3.83it/s]
 80% 60/75 [00:15<00:04,  3.61it/s]
 81% 61/75 [00:16<00:03,  3.59it/s]
 83% 62/75 [00:16<00:03,  3.46it/s]
 84% 63/75 [00:16<00:03,  3.46it/s]
 85% 64/75 [00:17<00:03,  3.51it/s]
 87% 65/75 [00:17<00:02,  3.55it/s]
 88% 66/75 [00:17<00:02,  3.97it/s]
 89% 67/75 [00:17<00:01,  4.18it/s]
 91% 68/75 [00:18<00:01,  3.92it/s]
 92% 69/75 [00:18<00:01,  3.87it/s]
 93% 70/75 [00:18<00:01,  4.18it/s]
 95% 71/75 [00:18<00:00,  4.08it/s]
 96% 72/75 [00:19<00:00,  3.92it/s]
 97% 73/75 [00:19<00:00,  4.13it/s]
 99% 74/75 [00:19<00:00,  3.80it/s]
                                         
{'eval_loss': 1.5661282539367676, 'eval_runtime': 20.0963, 'eval_samples_per_second': 3.732, 'eval_steps_per_second': 3.732, 'epoch': 0.0}
  0% 20/22194 [01:41<16:12:02,  2.63s/it]
100% 75/75 [00:20<00:00,  3.73it/s]
{'loss': 1.5859, 'learning_rate': 0.00019999987866801044, 'epoch': 0.0}
{'loss': 1.4043, 'learning_rate': 0.00019999985560491054, 'epoch': 0.0}
{'loss': 1.3492, 'learning_rate': 0.00019999983053632572, 'epoch': 0.0}
{'loss': 1.1044, 'learning_rate': 0.0001999998034622564, 'epoch': 0.0}
{'loss': 2.3291, 'learning_rate': 0.00019999977438270325, 'epoch': 0.0}
  0% 25/22194 [01:49<17:23:17,  2.82s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.02it/s]
  5% 4/75 [00:00<00:13,  5.17it/s]
  7% 5/75 [00:00<00:15,  4.55it/s]
  8% 6/75 [00:01<00:15,  4.42it/s]
  9% 7/75 [00:01<00:17,  3.86it/s]
 11% 8/75 [00:01<00:17,  3.75it/s]
 12% 9/75 [00:02<00:16,  3.95it/s]
 13% 10/75 [00:02<00:16,  3.97it/s]
 15% 11/75 [00:02<00:15,  4.10it/s]
 16% 12/75 [00:02<00:15,  3.99it/s]
 17% 13/75 [00:03<00:16,  3.83it/s]
 19% 14/75 [00:03<00:15,  3.83it/s]
 20% 15/75 [00:03<00:14,  4.07it/s]
 21% 16/75 [00:03<00:15,  3.74it/s]
 23% 17/75 [00:04<00:15,  3.82it/s]
 24% 18/75 [00:04<00:13,  4.09it/s]
 25% 19/75 [00:04<00:14,  3.90it/s]
 27% 20/75 [00:04<00:14,  3.69it/s]
 28% 21/75 [00:05<00:14,  3.75it/s]
 29% 22/75 [00:05<00:14,  3.68it/s]
 31% 23/75 [00:05<00:14,  3.47it/s]
 32% 24/75 [00:06<00:14,  3.50it/s]
 33% 25/75 [00:06<00:13,  3.84it/s]
 35% 26/75 [00:06<00:11,  4.16it/s]
 36% 27/75 [00:06<00:12,  3.95it/s]
 37% 28/75 [00:07<00:12,  3.84it/s]
 39% 29/75 [00:07<00:12,  3.79it/s]
 40% 30/75 [00:07<00:12,  3.57it/s]
 41% 31/75 [00:07<00:12,  3.44it/s]
 43% 32/75 [00:08<00:11,  3.75it/s]
 44% 33/75 [00:08<00:11,  3.74it/s]
 45% 34/75 [00:08<00:11,  3.67it/s]
 47% 35/75 [00:09<00:11,  3.50it/s]
 48% 36/75 [00:09<00:11,  3.54it/s]
 49% 37/75 [00:09<00:10,  3.53it/s]
 51% 38/75 [00:09<00:10,  3.47it/s]
 52% 39/75 [00:10<00:10,  3.47it/s]
 53% 40/75 [00:10<00:09,  3.63it/s]
 55% 41/75 [00:10<00:09,  3.56it/s]
 56% 42/75 [00:10<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.61it/s]
 59% 44/75 [00:11<00:07,  3.93it/s]
 60% 45/75 [00:11<00:07,  3.80it/s]
 61% 46/75 [00:12<00:08,  3.48it/s]
 63% 47/75 [00:12<00:08,  3.33it/s]
 64% 48/75 [00:12<00:08,  3.26it/s]
 65% 49/75 [00:13<00:07,  3.37it/s]
 67% 50/75 [00:13<00:07,  3.27it/s]
 68% 51/75 [00:13<00:07,  3.37it/s]
 69% 52/75 [00:13<00:06,  3.63it/s]
 71% 53/75 [00:14<00:06,  3.19it/s]
 72% 54/75 [00:14<00:06,  3.12it/s]
 73% 55/75 [00:14<00:05,  3.55it/s]
 75% 56/75 [00:15<00:05,  3.48it/s]
 76% 57/75 [00:15<00:05,  3.43it/s]
 77% 58/75 [00:15<00:04,  3.60it/s]
 79% 59/75 [00:15<00:04,  3.75it/s]
 80% 60/75 [00:16<00:04,  3.59it/s]
 81% 61/75 [00:16<00:04,  3.48it/s]
 83% 62/75 [00:16<00:03,  3.38it/s]
 84% 63/75 [00:17<00:03,  3.42it/s]
 85% 64/75 [00:17<00:03,  3.48it/s]
 87% 65/75 [00:17<00:02,  3.61it/s]
 88% 66/75 [00:17<00:02,  3.95it/s]
 89% 67/75 [00:18<00:01,  4.06it/s]
 91% 68/75 [00:18<00:01,  3.87it/s]
 92% 69/75 [00:18<00:01,  3.87it/s]
 93% 70/75 [00:18<00:01,  4.07it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.91it/s]
 97% 73/75 [00:19<00:00,  3.93it/s]
 99% 74/75 [00:19<00:00,  3.70it/s]
                                         
{'eval_loss': 1.5116215944290161, 'eval_runtime': 20.4166, 'eval_samples_per_second': 3.673, 'eval_steps_per_second': 3.673, 'epoch': 0.0}
  0% 25/22194 [02:09<17:23:17,  2.82s/it]
100% 75/75 [00:20<00:00,  3.66it/s]
{'loss': 2.0938, 'learning_rate': 0.0001999997432976668, 'epoch': 0.0}
{'loss': 1.7133, 'learning_rate': 0.00019999971020714762, 'epoch': 0.0}
{'loss': 1.7011, 'learning_rate': 0.00019999967511114645, 'epoch': 0.0}
{'loss': 2.869, 'learning_rate': 0.00019999963800966395, 'epoch': 0.0}
{'loss': 1.6705, 'learning_rate': 0.00019999959890270086, 'epoch': 0.0}
  0% 30/22194 [02:14<16:21:34,  2.66s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:08,  9.05it/s]
  4% 3/75 [00:00<00:10,  6.91it/s]
  5% 4/75 [00:00<00:13,  5.08it/s]
  7% 5/75 [00:01<00:17,  4.11it/s]
  8% 6/75 [00:01<00:15,  4.34it/s]
  9% 7/75 [00:01<00:18,  3.74it/s]
 11% 8/75 [00:01<00:19,  3.52it/s]
 12% 9/75 [00:02<00:17,  3.83it/s]
 13% 10/75 [00:02<00:17,  3.78it/s]
 15% 11/75 [00:02<00:15,  4.05it/s]
 16% 12/75 [00:02<00:16,  3.91it/s]
 17% 13/75 [00:03<00:16,  3.83it/s]
 19% 14/75 [00:03<00:15,  3.85it/s]
 20% 15/75 [00:03<00:14,  4.04it/s]
 21% 16/75 [00:03<00:15,  3.76it/s]
 23% 17/75 [00:04<00:15,  3.80it/s]
 24% 18/75 [00:04<00:14,  4.01it/s]
 25% 19/75 [00:04<00:14,  3.81it/s]
 27% 20/75 [00:04<00:14,  3.72it/s]
 28% 21/75 [00:05<00:14,  3.67it/s]
 29% 22/75 [00:05<00:14,  3.61it/s]
 31% 23/75 [00:05<00:15,  3.45it/s]
 32% 24/75 [00:06<00:14,  3.57it/s]
 33% 25/75 [00:06<00:12,  3.87it/s]
 35% 26/75 [00:06<00:12,  4.04it/s]
 36% 27/75 [00:06<00:12,  3.86it/s]
 37% 28/75 [00:07<00:12,  3.77it/s]
 39% 29/75 [00:07<00:12,  3.69it/s]
 40% 30/75 [00:07<00:12,  3.49it/s]
 41% 31/75 [00:08<00:12,  3.48it/s]
 43% 32/75 [00:08<00:11,  3.71it/s]
 44% 33/75 [00:08<00:11,  3.69it/s]
 45% 34/75 [00:08<00:11,  3.55it/s]
 47% 35/75 [00:09<00:11,  3.48it/s]
 48% 36/75 [00:09<00:11,  3.51it/s]
 49% 37/75 [00:09<00:10,  3.52it/s]
 51% 38/75 [00:09<00:10,  3.53it/s]
 52% 39/75 [00:10<00:10,  3.53it/s]
 53% 40/75 [00:10<00:09,  3.56it/s]
 55% 41/75 [00:10<00:09,  3.58it/s]
 56% 42/75 [00:11<00:09,  3.57it/s]
 57% 43/75 [00:11<00:08,  3.67it/s]
 59% 44/75 [00:11<00:08,  3.85it/s]
 60% 45/75 [00:11<00:08,  3.69it/s]
 61% 46/75 [00:12<00:08,  3.46it/s]
 63% 47/75 [00:12<00:08,  3.30it/s]
 64% 48/75 [00:12<00:08,  3.30it/s]
 65% 49/75 [00:13<00:07,  3.35it/s]
 67% 50/75 [00:13<00:07,  3.31it/s]
 68% 51/75 [00:13<00:07,  3.39it/s]
 69% 52/75 [00:13<00:06,  3.60it/s]
 71% 53/75 [00:14<00:06,  3.20it/s]
 72% 54/75 [00:14<00:06,  3.13it/s]
 73% 55/75 [00:14<00:05,  3.53it/s]
 75% 56/75 [00:15<00:05,  3.47it/s]
 76% 57/75 [00:15<00:05,  3.42it/s]
 77% 58/75 [00:15<00:04,  3.49it/s]
 79% 59/75 [00:15<00:04,  3.79it/s]
 80% 60/75 [00:16<00:04,  3.57it/s]
 81% 61/75 [00:16<00:03,  3.54it/s]
 83% 62/75 [00:16<00:03,  3.35it/s]
 84% 63/75 [00:17<00:03,  3.37it/s]
 85% 64/75 [00:17<00:03,  3.49it/s]
 87% 65/75 [00:17<00:02,  3.53it/s]
 88% 66/75 [00:17<00:02,  3.88it/s]
 89% 67/75 [00:18<00:01,  4.10it/s]
 91% 68/75 [00:18<00:01,  3.91it/s]
 92% 69/75 [00:18<00:01,  3.85it/s]
 93% 70/75 [00:18<00:01,  4.17it/s]
 95% 71/75 [00:19<00:00,  4.04it/s]
 96% 72/75 [00:19<00:00,  3.87it/s]
 97% 73/75 [00:19<00:00,  4.14it/s]
 99% 74/75 [00:19<00:00,  3.84it/s]
                                         
{'eval_loss': 1.4590259790420532, 'eval_runtime': 20.5037, 'eval_samples_per_second': 3.658, 'eval_steps_per_second': 3.658, 'epoch': 0.0}
  0% 30/22194 [02:34<16:21:34,  2.66s/it]
100% 75/75 [00:20<00:00,  3.75it/s]
{'loss': 0.6917, 'learning_rate': 0.00019999955779025801, 'epoch': 0.0}
{'loss': 1.3734, 'learning_rate': 0.00019999951467233618, 'epoch': 0.0}
{'loss': 1.0721, 'learning_rate': 0.00019999946954893627, 'epoch': 0.0}
{'loss': 2.0529, 'learning_rate': 0.00019999942242005916, 'epoch': 0.0}
{'loss': 1.3288, 'learning_rate': 0.00019999937328570578, 'epoch': 0.0}
  0% 35/22194 [02:41<17:22:43,  2.82s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.33it/s]
  5% 4/75 [00:00<00:13,  5.23it/s]
  7% 5/75 [00:00<00:15,  4.56it/s]
  8% 6/75 [00:01<00:15,  4.49it/s]
  9% 7/75 [00:01<00:17,  3.89it/s]
 11% 8/75 [00:01<00:17,  3.78it/s]
 12% 9/75 [00:02<00:16,  3.94it/s]
 13% 10/75 [00:02<00:16,  3.83it/s]
 15% 11/75 [00:02<00:15,  4.21it/s]
 16% 12/75 [00:02<00:15,  4.08it/s]
 17% 13/75 [00:03<00:15,  3.99it/s]
 19% 14/75 [00:03<00:15,  3.86it/s]
 20% 15/75 [00:03<00:14,  4.12it/s]
 21% 16/75 [00:03<00:15,  3.83it/s]
 23% 17/75 [00:04<00:15,  3.82it/s]
 24% 18/75 [00:04<00:13,  4.10it/s]
 25% 19/75 [00:04<00:14,  3.90it/s]
 27% 20/75 [00:04<00:14,  3.70it/s]
 28% 21/75 [00:05<00:14,  3.70it/s]
 29% 22/75 [00:05<00:14,  3.73it/s]
 31% 23/75 [00:05<00:14,  3.50it/s]
 32% 24/75 [00:06<00:14,  3.57it/s]
 33% 25/75 [00:06<00:12,  3.90it/s]
 35% 26/75 [00:06<00:11,  4.19it/s]
 36% 27/75 [00:06<00:12,  3.97it/s]
 37% 28/75 [00:06<00:12,  3.88it/s]
 39% 29/75 [00:07<00:12,  3.81it/s]
 40% 30/75 [00:07<00:12,  3.60it/s]
 41% 31/75 [00:07<00:12,  3.48it/s]
 43% 32/75 [00:08<00:11,  3.82it/s]
 44% 33/75 [00:08<00:10,  3.83it/s]
 45% 34/75 [00:08<00:11,  3.68it/s]
 47% 35/75 [00:08<00:11,  3.57it/s]
 48% 36/75 [00:09<00:10,  3.58it/s]
 49% 37/75 [00:09<00:10,  3.60it/s]
 51% 38/75 [00:09<00:10,  3.61it/s]
 52% 39/75 [00:10<00:10,  3.59it/s]
 53% 40/75 [00:10<00:09,  3.63it/s]
 55% 41/75 [00:10<00:09,  3.64it/s]
 56% 42/75 [00:10<00:09,  3.63it/s]
 57% 43/75 [00:11<00:08,  3.75it/s]
 59% 44/75 [00:11<00:07,  3.93it/s]
 60% 45/75 [00:11<00:07,  3.77it/s]
 61% 46/75 [00:11<00:08,  3.52it/s]
 63% 47/75 [00:12<00:08,  3.35it/s]
 64% 48/75 [00:12<00:08,  3.34it/s]
 65% 49/75 [00:12<00:07,  3.38it/s]
 67% 50/75 [00:13<00:07,  3.34it/s]
 68% 51/75 [00:13<00:06,  3.49it/s]
 69% 52/75 [00:13<00:06,  3.57it/s]
 71% 53/75 [00:14<00:06,  3.18it/s]
 72% 54/75 [00:14<00:06,  3.23it/s]
 73% 55/75 [00:14<00:05,  3.51it/s]
 75% 56/75 [00:14<00:05,  3.43it/s]
 76% 57/75 [00:15<00:05,  3.40it/s]
 77% 58/75 [00:15<00:04,  3.57it/s]
 79% 59/75 [00:15<00:04,  3.73it/s]
 80% 60/75 [00:16<00:04,  3.58it/s]
 81% 61/75 [00:16<00:03,  3.50it/s]
 83% 62/75 [00:16<00:03,  3.39it/s]
 84% 63/75 [00:16<00:03,  3.44it/s]
 85% 64/75 [00:17<00:03,  3.50it/s]
 87% 65/75 [00:17<00:02,  3.63it/s]
 88% 66/75 [00:17<00:02,  3.98it/s]
 89% 67/75 [00:17<00:01,  4.09it/s]
 91% 68/75 [00:18<00:01,  3.92it/s]
 92% 69/75 [00:18<00:01,  3.83it/s]
 93% 70/75 [00:18<00:01,  4.12it/s]
 95% 71/75 [00:18<00:01,  3.99it/s]
 96% 72/75 [00:19<00:00,  3.85it/s]
 97% 73/75 [00:19<00:00,  4.02it/s]
 99% 74/75 [00:19<00:00,  3.75it/s]
                                         
{'eval_loss': 1.4234569072723389, 'eval_runtime': 20.2373, 'eval_samples_per_second': 3.706, 'eval_steps_per_second': 3.706, 'epoch': 0.0}
  0% 35/22194 [03:01<17:22:43,  2.82s/it]
100% 75/75 [00:20<00:00,  3.71it/s]
{'loss': 2.1383, 'learning_rate': 0.00019999932214587713, 'epoch': 0.0}
{'loss': 0.9362, 'learning_rate': 0.00019999926900057425, 'epoch': 0.01}
{'loss': 1.3196, 'learning_rate': 0.00019999921384979821, 'epoch': 0.01}
{'loss': 0.9733, 'learning_rate': 0.00019999915669355008, 'epoch': 0.01}
{'loss': 2.0384, 'learning_rate': 0.00019999909753183105, 'epoch': 0.01}
  0% 40/22194 [03:06<15:49:33,  2.57s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.84it/s]
  4% 3/75 [00:00<00:10,  6.68it/s]
  5% 4/75 [00:00<00:14,  5.03it/s]
  7% 5/75 [00:01<00:16,  4.24it/s]
  8% 6/75 [00:01<00:16,  4.24it/s]
  9% 7/75 [00:01<00:18,  3.77it/s]
 11% 8/75 [00:01<00:18,  3.64it/s]
 12% 9/75 [00:02<00:17,  3.83it/s]
 13% 10/75 [00:02<00:16,  3.92it/s]
 15% 11/75 [00:02<00:15,  4.12it/s]
 16% 12/75 [00:02<00:15,  3.95it/s]
 17% 13/75 [00:03<00:16,  3.84it/s]
 19% 14/75 [00:03<00:15,  3.88it/s]
 20% 15/75 [00:03<00:15,  3.97it/s]
 21% 16/75 [00:03<00:15,  3.69it/s]
 23% 17/75 [00:04<00:15,  3.80it/s]
 24% 18/75 [00:04<00:14,  3.99it/s]
 25% 19/75 [00:04<00:14,  3.82it/s]
 27% 20/75 [00:04<00:14,  3.68it/s]
 28% 21/75 [00:05<00:14,  3.65it/s]
 29% 22/75 [00:05<00:14,  3.60it/s]
 31% 23/75 [00:05<00:15,  3.44it/s]
 32% 24/75 [00:06<00:14,  3.56it/s]
 33% 25/75 [00:06<00:12,  3.88it/s]
 35% 26/75 [00:06<00:12,  4.03it/s]
 36% 27/75 [00:06<00:12,  3.89it/s]
 37% 28/75 [00:07<00:12,  3.75it/s]
 39% 29/75 [00:07<00:12,  3.69it/s]
 40% 30/75 [00:07<00:12,  3.48it/s]
 41% 31/75 [00:08<00:12,  3.45it/s]
 43% 32/75 [00:08<00:11,  3.69it/s]
 44% 33/75 [00:08<00:11,  3.69it/s]
 45% 34/75 [00:08<00:11,  3.57it/s]
 47% 35/75 [00:09<00:11,  3.48it/s]
 48% 36/75 [00:09<00:11,  3.50it/s]
 49% 37/75 [00:09<00:10,  3.51it/s]
 51% 38/75 [00:09<00:10,  3.51it/s]
 52% 39/75 [00:10<00:10,  3.46it/s]
 53% 40/75 [00:10<00:09,  3.54it/s]
 55% 41/75 [00:10<00:09,  3.58it/s]
 56% 42/75 [00:11<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.64it/s]
 59% 44/75 [00:11<00:07,  3.89it/s]
 60% 45/75 [00:11<00:08,  3.74it/s]
 61% 46/75 [00:12<00:08,  3.48it/s]
 63% 47/75 [00:12<00:08,  3.17it/s]
 64% 48/75 [00:12<00:08,  3.26it/s]
 65% 49/75 [00:13<00:07,  3.34it/s]
 67% 50/75 [00:13<00:07,  3.24it/s]
 68% 51/75 [00:13<00:07,  3.35it/s]
 69% 52/75 [00:13<00:06,  3.62it/s]
 71% 53/75 [00:14<00:06,  3.20it/s]
 72% 54/75 [00:14<00:06,  3.12it/s]
 73% 55/75 [00:14<00:05,  3.50it/s]
 75% 56/75 [00:15<00:05,  3.49it/s]
 76% 57/75 [00:15<00:05,  3.39it/s]
 77% 58/75 [00:15<00:04,  3.50it/s]
 79% 59/75 [00:15<00:04,  3.82it/s]
 80% 60/75 [00:16<00:04,  3.64it/s]
 81% 61/75 [00:16<00:03,  3.52it/s]
 83% 62/75 [00:16<00:03,  3.39it/s]
 84% 63/75 [00:17<00:03,  3.43it/s]
 85% 64/75 [00:17<00:03,  3.51it/s]
 87% 65/75 [00:17<00:02,  3.65it/s]
 88% 66/75 [00:17<00:02,  3.97it/s]
 89% 67/75 [00:18<00:01,  4.06it/s]
 91% 68/75 [00:18<00:01,  3.90it/s]
 92% 69/75 [00:18<00:01,  3.90it/s]
 93% 70/75 [00:18<00:01,  4.09it/s]
 95% 71/75 [00:19<00:01,  3.94it/s]
 96% 72/75 [00:19<00:00,  3.91it/s]
 97% 73/75 [00:19<00:00,  3.97it/s]
 99% 74/75 [00:19<00:00,  3.73it/s]
                                         
{'eval_loss': 1.394790530204773, 'eval_runtime': 20.5085, 'eval_samples_per_second': 3.657, 'eval_steps_per_second': 3.657, 'epoch': 0.01}
  0% 40/22194 [03:26<15:49:33,  2.57s/it]
100% 75/75 [00:20<00:00,  3.68it/s]
{'loss': 1.4345, 'learning_rate': 0.00019999903636464233, 'epoch': 0.01}
{'loss': 1.1868, 'learning_rate': 0.00019999897319198504, 'epoch': 0.01}
{'loss': 1.4663, 'learning_rate': 0.00019999890801386053, 'epoch': 0.01}
{'loss': 0.9754, 'learning_rate': 0.00019999884083027008, 'epoch': 0.01}
{'loss': 1.6294, 'learning_rate': 0.00019999877164121504, 'epoch': 0.01}
  0% 45/22194 [03:33<17:11:14,  2.79s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.02it/s]
  5% 4/75 [00:00<00:13,  5.38it/s]
  7% 5/75 [00:00<00:15,  4.53it/s]
  8% 6/75 [00:01<00:15,  4.55it/s]
  9% 7/75 [00:01<00:17,  3.92it/s]
 11% 8/75 [00:01<00:18,  3.71it/s]
 12% 9/75 [00:02<00:16,  3.99it/s]
 13% 10/75 [00:02<00:16,  3.86it/s]
 15% 11/75 [00:02<00:15,  4.12it/s]
 16% 12/75 [00:02<00:15,  4.08it/s]
 17% 13/75 [00:03<00:15,  3.92it/s]
 19% 14/75 [00:03<00:15,  3.86it/s]
 20% 15/75 [00:03<00:14,  4.12it/s]
 21% 16/75 [00:03<00:15,  3.76it/s]
 23% 17/75 [00:04<00:15,  3.82it/s]
 24% 18/75 [00:04<00:14,  4.02it/s]
 25% 19/75 [00:04<00:14,  3.85it/s]
 27% 20/75 [00:04<00:14,  3.83it/s]
 28% 21/75 [00:05<00:14,  3.75it/s]
 29% 22/75 [00:05<00:14,  3.66it/s]
 31% 23/75 [00:05<00:14,  3.53it/s]
 32% 24/75 [00:06<00:13,  3.66it/s]
 33% 25/75 [00:06<00:12,  3.92it/s]
 35% 26/75 [00:06<00:12,  4.07it/s]
 36% 27/75 [00:06<00:12,  3.94it/s]
 37% 28/75 [00:07<00:12,  3.81it/s]
 39% 29/75 [00:07<00:12,  3.72it/s]
 40% 30/75 [00:07<00:12,  3.52it/s]
 41% 31/75 [00:07<00:12,  3.50it/s]
 43% 32/75 [00:08<00:11,  3.80it/s]
 44% 33/75 [00:08<00:11,  3.75it/s]
 45% 34/75 [00:08<00:11,  3.63it/s]
 47% 35/75 [00:08<00:11,  3.52it/s]
 48% 36/75 [00:09<00:11,  3.52it/s]
 49% 37/75 [00:09<00:10,  3.57it/s]
 51% 38/75 [00:09<00:10,  3.59it/s]
 52% 39/75 [00:10<00:10,  3.57it/s]
 53% 40/75 [00:10<00:09,  3.62it/s]
 55% 41/75 [00:10<00:09,  3.62it/s]
 56% 42/75 [00:10<00:09,  3.60it/s]
 57% 43/75 [00:11<00:08,  3.73it/s]
 59% 44/75 [00:11<00:07,  3.92it/s]
 60% 45/75 [00:11<00:07,  3.75it/s]
 61% 46/75 [00:12<00:08,  3.50it/s]
 63% 47/75 [00:12<00:08,  3.32it/s]
 64% 48/75 [00:12<00:08,  3.32it/s]
 65% 49/75 [00:12<00:07,  3.36it/s]
 67% 50/75 [00:13<00:07,  3.32it/s]
 68% 51/75 [00:13<00:06,  3.50it/s]
 69% 52/75 [00:13<00:06,  3.56it/s]
 71% 53/75 [00:14<00:06,  3.17it/s]
 72% 54/75 [00:14<00:06,  3.22it/s]
 73% 55/75 [00:14<00:05,  3.52it/s]
 75% 56/75 [00:14<00:05,  3.46it/s]
 76% 57/75 [00:15<00:05,  3.37it/s]
 77% 58/75 [00:15<00:04,  3.52it/s]
 79% 59/75 [00:15<00:04,  3.82it/s]
 80% 60/75 [00:16<00:04,  3.63it/s]
 81% 61/75 [00:16<00:03,  3.53it/s]
 83% 62/75 [00:16<00:03,  3.43it/s]
 84% 63/75 [00:16<00:03,  3.44it/s]
 85% 64/75 [00:17<00:03,  3.52it/s]
 87% 65/75 [00:17<00:02,  3.49it/s]
 88% 66/75 [00:17<00:02,  3.93it/s]
 89% 67/75 [00:17<00:01,  4.12it/s]
 91% 68/75 [00:18<00:01,  3.92it/s]
 92% 69/75 [00:18<00:01,  3.85it/s]
 93% 70/75 [00:18<00:01,  4.12it/s]
 95% 71/75 [00:18<00:01,  3.96it/s]
 96% 72/75 [00:19<00:00,  3.86it/s]
 97% 73/75 [00:19<00:00,  4.12it/s]
 99% 74/75 [00:19<00:00,  3.76it/s]
                                         
{'eval_loss': 1.4229274988174438, 'eval_runtime': 20.3105, 'eval_samples_per_second': 3.693, 'eval_steps_per_second': 3.693, 'epoch': 0.01}
  0% 45/22194 [03:53<17:11:14,  2.79s/it]
100% 75/75 [00:20<00:00,  3.69it/s]
{'loss': 1.4568, 'learning_rate': 0.0001999987004466968, 'epoch': 0.01}
{'loss': 1.2036, 'learning_rate': 0.00019999862724671678, 'epoch': 0.01}
{'loss': 1.2342, 'learning_rate': 0.0001999985520412765, 'epoch': 0.01}
{'loss': 1.374, 'learning_rate': 0.0001999984748303774, 'epoch': 0.01}
{'loss': 1.0639, 'learning_rate': 0.00019999839561402104, 'epoch': 0.01}
  0% 50/22194 [03:57<15:19:05,  2.49s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.70it/s]
  5% 4/75 [00:00<00:13,  5.25it/s]
  7% 5/75 [00:00<00:15,  4.52it/s]
  8% 6/75 [00:01<00:15,  4.42it/s]
  9% 7/75 [00:01<00:17,  3.88it/s]
 11% 8/75 [00:01<00:17,  3.74it/s]
 12% 9/75 [00:02<00:16,  3.94it/s]
 13% 10/75 [00:02<00:16,  3.96it/s]
 15% 11/75 [00:02<00:15,  4.11it/s]
 16% 12/75 [00:02<00:15,  3.96it/s]
 17% 13/75 [00:03<00:16,  3.86it/s]
 19% 14/75 [00:03<00:15,  3.93it/s]
 20% 15/75 [00:03<00:15,  3.98it/s]
 21% 16/75 [00:03<00:15,  3.73it/s]
 23% 17/75 [00:04<00:15,  3.83it/s]
 24% 18/75 [00:04<00:14,  4.00it/s]
 25% 19/75 [00:04<00:14,  3.83it/s]
 27% 20/75 [00:04<00:14,  3.69it/s]
 28% 21/75 [00:05<00:14,  3.70it/s]
 29% 22/75 [00:05<00:14,  3.65it/s]
 31% 23/75 [00:05<00:14,  3.48it/s]
 32% 24/75 [00:06<00:14,  3.60it/s]
 33% 25/75 [00:06<00:12,  3.90it/s]
 35% 26/75 [00:06<00:12,  4.06it/s]
 36% 27/75 [00:06<00:12,  3.92it/s]
 37% 28/75 [00:07<00:12,  3.80it/s]
 39% 29/75 [00:07<00:12,  3.71it/s]
 40% 30/75 [00:07<00:12,  3.50it/s]
 41% 31/75 [00:07<00:12,  3.40it/s]
 43% 32/75 [00:08<00:11,  3.78it/s]
 44% 33/75 [00:08<00:11,  3.76it/s]
 45% 34/75 [00:08<00:11,  3.61it/s]
 47% 35/75 [00:09<00:11,  3.46it/s]
 48% 36/75 [00:09<00:11,  3.54it/s]
 49% 37/75 [00:09<00:10,  3.55it/s]
 51% 38/75 [00:09<00:10,  3.56it/s]
 52% 39/75 [00:10<00:10,  3.50it/s]
 53% 40/75 [00:10<00:09,  3.58it/s]
 55% 41/75 [00:10<00:09,  3.60it/s]
 56% 42/75 [00:11<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.59it/s]
 59% 44/75 [00:11<00:07,  3.89it/s]
 60% 45/75 [00:11<00:07,  3.77it/s]
 61% 46/75 [00:12<00:08,  3.50it/s]
 63% 47/75 [00:12<00:08,  3.33it/s]
 64% 48/75 [00:12<00:08,  3.24it/s]
 65% 49/75 [00:13<00:07,  3.36it/s]
 67% 50/75 [00:13<00:07,  3.25it/s]
 68% 51/75 [00:13<00:06,  3.50it/s]
 69% 52/75 [00:13<00:06,  3.57it/s]
 71% 53/75 [00:14<00:06,  3.17it/s]
 72% 54/75 [00:14<00:06,  3.21it/s]
 73% 55/75 [00:14<00:05,  3.49it/s]
 75% 56/75 [00:15<00:05,  3.47it/s]
 76% 57/75 [00:15<00:05,  3.42it/s]
 77% 58/75 [00:15<00:04,  3.57it/s]
 79% 59/75 [00:15<00:04,  3.74it/s]
 80% 60/75 [00:16<00:04,  3.59it/s]
 81% 61/75 [00:16<00:04,  3.50it/s]
 83% 62/75 [00:16<00:03,  3.38it/s]
 84% 63/75 [00:17<00:03,  3.40it/s]
 85% 64/75 [00:17<00:03,  3.49it/s]
 87% 65/75 [00:17<00:02,  3.62it/s]
 88% 66/75 [00:17<00:02,  3.95it/s]
 89% 67/75 [00:18<00:01,  4.06it/s]
 91% 68/75 [00:18<00:01,  3.89it/s]
 92% 69/75 [00:18<00:01,  3.89it/s]
 93% 70/75 [00:18<00:01,  4.08it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.91it/s]
 97% 73/75 [00:19<00:00,  3.98it/s]
 99% 74/75 [00:19<00:00,  3.75it/s]
                                         
{'eval_loss': 1.379117488861084, 'eval_runtime': 20.4253, 'eval_samples_per_second': 3.672, 'eval_steps_per_second': 3.672, 'epoch': 0.01}
  0% 50/22194 [04:17<15:19:05,  2.49s/it]
100% 75/75 [00:20<00:00,  3.69it/s]
{'loss': 0.8922, 'learning_rate': 0.00019999831439220905, 'epoch': 0.01}
{'loss': 1.7158, 'learning_rate': 0.00019999823116494301, 'epoch': 0.01}
{'loss': 1.4994, 'learning_rate': 0.00019999814593222462, 'epoch': 0.01}
{'loss': 1.5456, 'learning_rate': 0.00019999805869405556, 'epoch': 0.01}
{'loss': 2.0237, 'learning_rate': 0.00019999796945043764, 'epoch': 0.01}
  0% 55/22194 [04:25<18:00:06,  2.93s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.80it/s]
  5% 4/75 [00:00<00:12,  5.84it/s]
  7% 5/75 [00:00<00:15,  4.44it/s]
  8% 6/75 [00:01<00:14,  4.63it/s]
  9% 7/75 [00:01<00:16,  4.07it/s]
 11% 8/75 [00:01<00:17,  3.82it/s]
 12% 9/75 [00:02<00:16,  3.98it/s]
 13% 10/75 [00:02<00:16,  4.03it/s]
 15% 11/75 [00:02<00:15,  4.19it/s]
 16% 12/75 [00:02<00:15,  4.00it/s]
 17% 13/75 [00:03<00:16,  3.87it/s]
 19% 14/75 [00:03<00:15,  3.98it/s]
 20% 15/75 [00:03<00:14,  4.07it/s]
 21% 16/75 [00:03<00:15,  3.73it/s]
 23% 17/75 [00:04<00:15,  3.79it/s]
 24% 18/75 [00:04<00:14,  4.03it/s]
 25% 19/75 [00:04<00:14,  3.85it/s]
 27% 20/75 [00:04<00:14,  3.71it/s]
 28% 21/75 [00:05<00:14,  3.69it/s]
 29% 22/75 [00:05<00:14,  3.65it/s]
 31% 23/75 [00:05<00:14,  3.48it/s]
 32% 24/75 [00:06<00:14,  3.59it/s]
 33% 25/75 [00:06<00:12,  3.91it/s]
 35% 26/75 [00:06<00:11,  4.08it/s]
 36% 27/75 [00:06<00:12,  3.93it/s]
 37% 28/75 [00:06<00:12,  3.76it/s]
 39% 29/75 [00:07<00:12,  3.70it/s]
 40% 30/75 [00:07<00:12,  3.51it/s]
 41% 31/75 [00:07<00:12,  3.47it/s]
 43% 32/75 [00:08<00:11,  3.73it/s]
 44% 33/75 [00:08<00:11,  3.73it/s]
 45% 34/75 [00:08<00:11,  3.60it/s]
 47% 35/75 [00:08<00:11,  3.48it/s]
 48% 36/75 [00:09<00:11,  3.51it/s]
 49% 37/75 [00:09<00:10,  3.54it/s]
 51% 38/75 [00:09<00:10,  3.52it/s]
 52% 39/75 [00:10<00:10,  3.54it/s]
 53% 40/75 [00:10<00:09,  3.59it/s]
 55% 41/75 [00:10<00:09,  3.59it/s]
 56% 42/75 [00:10<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.65it/s]
 59% 44/75 [00:11<00:08,  3.85it/s]
 60% 45/75 [00:11<00:08,  3.71it/s]
 61% 46/75 [00:12<00:08,  3.45it/s]
 63% 47/75 [00:12<00:08,  3.31it/s]
 64% 48/75 [00:12<00:08,  3.30it/s]
 65% 49/75 [00:12<00:07,  3.32it/s]
 67% 50/75 [00:13<00:07,  3.26it/s]
 68% 51/75 [00:13<00:07,  3.38it/s]
 69% 52/75 [00:13<00:06,  3.57it/s]
 71% 53/75 [00:14<00:06,  3.15it/s]
 72% 54/75 [00:14<00:06,  3.12it/s]
 73% 55/75 [00:14<00:05,  3.48it/s]
 75% 56/75 [00:15<00:05,  3.43it/s]
 76% 57/75 [00:15<00:05,  3.38it/s]
 77% 58/75 [00:15<00:04,  3.49it/s]
 79% 59/75 [00:15<00:04,  3.82it/s]
 80% 60/75 [00:16<00:04,  3.56it/s]
 81% 61/75 [00:16<00:03,  3.53it/s]
 83% 62/75 [00:16<00:03,  3.34it/s]
 84% 63/75 [00:17<00:03,  3.36it/s]
 85% 64/75 [00:17<00:03,  3.45it/s]
 87% 65/75 [00:17<00:02,  3.54it/s]
 88% 66/75 [00:17<00:02,  3.87it/s]
 89% 67/75 [00:18<00:01,  4.15it/s]
 91% 68/75 [00:18<00:01,  3.94it/s]
 92% 69/75 [00:18<00:01,  3.92it/s]
 93% 70/75 [00:18<00:01,  4.09it/s]
 95% 71/75 [00:19<00:01,  3.96it/s]
 96% 72/75 [00:19<00:00,  3.93it/s]
 97% 73/75 [00:19<00:00,  3.96it/s]
 99% 74/75 [00:19<00:00,  3.73it/s]
                                         
{'eval_loss': 1.3518704175949097, 'eval_runtime': 20.4505, 'eval_samples_per_second': 3.667, 'eval_steps_per_second': 3.667, 'epoch': 0.01}
  0% 55/22194 [04:45<18:00:06,  2.93s/it]
100% 75/75 [00:20<00:00,  3.69it/s]
{'loss': 0.8913, 'learning_rate': 0.0001999978782013726, 'epoch': 0.01}
{'loss': 1.2633, 'learning_rate': 0.00019999778494686225, 'epoch': 0.01}
{'loss': 1.0411, 'learning_rate': 0.0001999976896869085, 'epoch': 0.01}
{'loss': 1.7025, 'learning_rate': 0.00019999759242151326, 'epoch': 0.01}
{'loss': 0.5894, 'learning_rate': 0.00019999749315067846, 'epoch': 0.01}
  0% 60/22194 [04:49<16:44:57,  2.72s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 11.06it/s]
  5% 4/75 [00:00<00:13,  5.37it/s]
  7% 5/75 [00:01<00:16,  4.28it/s]
  8% 6/75 [00:01<00:15,  4.52it/s]
  9% 7/75 [00:01<00:17,  3.86it/s]
 11% 8/75 [00:01<00:18,  3.66it/s]
 12% 9/75 [00:02<00:16,  3.99it/s]
 13% 10/75 [00:02<00:16,  4.00it/s]
 15% 11/75 [00:02<00:15,  4.14it/s]
 16% 12/75 [00:02<00:15,  3.95it/s]
 17% 13/75 [00:03<00:16,  3.83it/s]
 19% 14/75 [00:03<00:15,  3.82it/s]
 20% 15/75 [00:03<00:14,  4.01it/s]
 21% 16/75 [00:03<00:15,  3.71it/s]
 23% 17/75 [00:04<00:15,  3.71it/s]
 24% 18/75 [00:04<00:14,  3.98it/s]
 25% 19/75 [00:04<00:14,  3.78it/s]
 27% 20/75 [00:04<00:14,  3.67it/s]
 28% 21/75 [00:05<00:14,  3.70it/s]
 29% 22/75 [00:05<00:14,  3.66it/s]
 31% 23/75 [00:05<00:15,  3.40it/s]
 32% 24/75 [00:06<00:14,  3.46it/s]
 33% 25/75 [00:06<00:12,  3.86it/s]
 35% 26/75 [00:06<00:11,  4.09it/s]
 36% 27/75 [00:06<00:12,  3.92it/s]
 37% 28/75 [00:07<00:12,  3.73it/s]
 39% 29/75 [00:07<00:12,  3.78it/s]
 40% 30/75 [00:07<00:13,  3.45it/s]
 41% 31/75 [00:08<00:12,  3.44it/s]
 43% 32/75 [00:08<00:11,  3.78it/s]
 44% 33/75 [00:08<00:11,  3.77it/s]
 45% 34/75 [00:08<00:11,  3.60it/s]
 47% 35/75 [00:09<00:11,  3.50it/s]
 48% 36/75 [00:09<00:11,  3.52it/s]
 49% 37/75 [00:09<00:10,  3.61it/s]
 51% 38/75 [00:09<00:10,  3.62it/s]
 52% 39/75 [00:10<00:10,  3.55it/s]
 53% 40/75 [00:10<00:10,  3.31it/s]
 55% 41/75 [00:10<00:10,  3.26it/s]
 56% 42/75 [00:11<00:08,  3.69it/s]
 57% 43/75 [00:11<00:08,  3.72it/s]
 59% 44/75 [00:11<00:08,  3.76it/s]
 60% 45/75 [00:11<00:07,  3.75it/s]
 61% 46/75 [00:12<00:08,  3.41it/s]
 63% 47/75 [00:12<00:08,  3.44it/s]
 64% 48/75 [00:12<00:08,  3.30it/s]
 65% 49/75 [00:13<00:07,  3.46it/s]
 67% 50/75 [00:13<00:08,  2.99it/s]
 68% 51/75 [00:13<00:07,  3.42it/s]
 69% 52/75 [00:13<00:06,  3.76it/s]
 71% 53/75 [00:14<00:06,  3.25it/s]
 72% 54/75 [00:14<00:06,  3.25it/s]
 73% 55/75 [00:14<00:05,  3.57it/s]
 75% 56/75 [00:15<00:05,  3.54it/s]
 76% 57/75 [00:15<00:05,  3.46it/s]
 77% 58/75 [00:15<00:04,  3.58it/s]
 79% 59/75 [00:15<00:04,  3.82it/s]
 80% 60/75 [00:16<00:04,  3.64it/s]
 81% 61/75 [00:16<00:04,  3.49it/s]
 83% 62/75 [00:16<00:03,  3.40it/s]
 84% 63/75 [00:17<00:03,  3.44it/s]
 85% 64/75 [00:17<00:03,  3.48it/s]
 87% 65/75 [00:17<00:02,  3.60it/s]
 88% 66/75 [00:17<00:02,  3.95it/s]
 89% 67/75 [00:18<00:01,  4.10it/s]
 91% 68/75 [00:18<00:01,  3.90it/s]
 92% 69/75 [00:18<00:01,  3.89it/s]
 93% 70/75 [00:18<00:01,  4.11it/s]
 95% 71/75 [00:19<00:01,  3.99it/s]
 96% 72/75 [00:19<00:00,  3.93it/s]
 97% 73/75 [00:19<00:00,  3.99it/s]
 99% 74/75 [00:19<00:00,  3.78it/s]
                                         
{'eval_loss': 1.351861834526062, 'eval_runtime': 20.4514, 'eval_samples_per_second': 3.667, 'eval_steps_per_second': 3.667, 'epoch': 0.01}
  0% 60/22194 [05:10<16:44:57,  2.72s/it]
100% 75/75 [00:20<00:00,  3.71it/s]
{'loss': 0.8559, 'learning_rate': 0.00019999739187440612, 'epoch': 0.01}
{'loss': 1.1363, 'learning_rate': 0.00019999728859269826, 'epoch': 0.01}
{'loss': 1.3649, 'learning_rate': 0.0001999971833055569, 'epoch': 0.01}
{'loss': 1.9424, 'learning_rate': 0.00019999707601298423, 'epoch': 0.01}
{'loss': 1.5771, 'learning_rate': 0.00019999696671498237, 'epoch': 0.01}
  0% 65/22194 [05:26<21:09:35,  3.44s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.65it/s]
  4% 3/75 [00:00<00:09,  7.21it/s]
  5% 4/75 [00:00<00:13,  5.39it/s]
  7% 5/75 [00:00<00:16,  4.23it/s]
  8% 6/75 [00:01<00:15,  4.47it/s]
  9% 7/75 [00:01<00:17,  3.92it/s]
 11% 8/75 [00:01<00:18,  3.70it/s]
 12% 9/75 [00:02<00:16,  3.98it/s]
 13% 10/75 [00:02<00:16,  3.84it/s]
 15% 11/75 [00:02<00:15,  4.18it/s]
 16% 12/75 [00:02<00:15,  4.11it/s]
 17% 13/75 [00:03<00:15,  3.96it/s]
 19% 14/75 [00:03<00:15,  3.87it/s]
 20% 15/75 [00:03<00:14,  4.20it/s]
 21% 16/75 [00:03<00:15,  3.83it/s]
 23% 17/75 [00:04<00:15,  3.75it/s]
 24% 18/75 [00:04<00:13,  4.17it/s]
 25% 19/75 [00:04<00:13,  4.00it/s]
 27% 20/75 [00:04<00:14,  3.86it/s]
 28% 21/75 [00:05<00:14,  3.78it/s]
 29% 22/75 [00:05<00:14,  3.69it/s]
 31% 23/75 [00:05<00:14,  3.56it/s]
 32% 24/75 [00:05<00:13,  3.67it/s]
 33% 25/75 [00:06<00:12,  3.98it/s]
 35% 26/75 [00:06<00:11,  4.16it/s]
 36% 27/75 [00:06<00:11,  4.03it/s]
 37% 28/75 [00:06<00:12,  3.87it/s]
 39% 29/75 [00:07<00:12,  3.76it/s]
 40% 30/75 [00:07<00:12,  3.56it/s]
 41% 31/75 [00:07<00:12,  3.54it/s]
 43% 32/75 [00:08<00:11,  3.61it/s]
 44% 33/75 [00:08<00:11,  3.64it/s]
 45% 34/75 [00:08<00:11,  3.66it/s]
 47% 35/75 [00:09<00:12,  3.18it/s]
 48% 36/75 [00:09<00:10,  3.61it/s]
 49% 37/75 [00:09<00:11,  3.35it/s]
 51% 38/75 [00:09<00:12,  2.98it/s]
 52% 39/75 [00:10<00:11,  3.22it/s]
 53% 40/75 [00:10<00:11,  3.18it/s]
 55% 41/75 [00:10<00:09,  3.72it/s]
 56% 42/75 [00:10<00:08,  3.68it/s]
 57% 43/75 [00:11<00:08,  3.70it/s]
 59% 44/75 [00:11<00:08,  3.69it/s]
 60% 45/75 [00:11<00:07,  3.76it/s]
 61% 46/75 [00:12<00:08,  3.54it/s]
 63% 47/75 [00:12<00:08,  3.23it/s]
 64% 48/75 [00:12<00:08,  3.20it/s]
 65% 49/75 [00:13<00:07,  3.37it/s]
 67% 50/75 [00:13<00:07,  3.19it/s]
 68% 51/75 [00:13<00:07,  3.28it/s]
 69% 52/75 [00:13<00:06,  3.55it/s]
 71% 53/75 [00:14<00:07,  3.12it/s]
 72% 54/75 [00:14<00:07,  2.97it/s]
 73% 55/75 [00:14<00:05,  3.41it/s]
 75% 56/75 [00:15<00:05,  3.43it/s]
 76% 57/75 [00:15<00:05,  3.27it/s]
 77% 58/75 [00:15<00:05,  3.39it/s]
 79% 59/75 [00:15<00:04,  3.72it/s]
 80% 60/75 [00:16<00:04,  3.43it/s]
 81% 61/75 [00:16<00:04,  3.41it/s]
 83% 62/75 [00:16<00:03,  3.26it/s]
 84% 63/75 [00:17<00:03,  3.29it/s]
 85% 64/75 [00:17<00:03,  3.33it/s]
 87% 65/75 [00:17<00:03,  3.33it/s]
 88% 66/75 [00:18<00:02,  3.70it/s]
 89% 67/75 [00:18<00:01,  4.02it/s]
 91% 68/75 [00:18<00:01,  3.70it/s]
 92% 69/75 [00:18<00:01,  3.56it/s]
 93% 70/75 [00:19<00:01,  3.98it/s]
 95% 71/75 [00:19<00:01,  3.83it/s]
 96% 72/75 [00:19<00:00,  3.64it/s]
 97% 73/75 [00:19<00:00,  3.87it/s]
 99% 74/75 [00:20<00:00,  3.57it/s]
                                         
{'eval_loss': 1.3509403467178345, 'eval_runtime': 20.793, 'eval_samples_per_second': 3.607, 'eval_steps_per_second': 3.607, 'epoch': 0.01}
  0% 65/22194 [05:47<21:09:35,  3.44s/it]
100% 75/75 [00:20<00:00,  3.45it/s]
{'loss': 1.1616, 'learning_rate': 0.0001999968554115535, 'epoch': 0.01}
{'loss': 2.0497, 'learning_rate': 0.00019999674210269985, 'epoch': 0.01}
{'loss': 1.7551, 'learning_rate': 0.00019999662678842374, 'epoch': 0.01}
{'loss': 0.6971, 'learning_rate': 0.00019999650946872738, 'epoch': 0.01}
{'loss': 1.427, 'learning_rate': 0.00019999639014361324, 'epoch': 0.01}
  0% 70/22194 [05:51<17:23:21,  2.83s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:08,  8.88it/s]
  4% 3/75 [00:00<00:11,  6.06it/s]
  5% 4/75 [00:00<00:15,  4.67it/s]
  7% 5/75 [00:01<00:17,  4.08it/s]
  8% 6/75 [00:01<00:17,  3.94it/s]
  9% 7/75 [00:01<00:21,  3.18it/s]
 11% 8/75 [00:02<00:20,  3.22it/s]
 12% 9/75 [00:02<00:19,  3.44it/s]
 13% 10/75 [00:02<00:18,  3.42it/s]
 15% 11/75 [00:02<00:16,  3.91it/s]
 16% 12/75 [00:03<00:15,  4.00it/s]
 17% 13/75 [00:03<00:16,  3.86it/s]
 19% 14/75 [00:03<00:16,  3.73it/s]
 20% 15/75 [00:04<00:19,  3.13it/s]
 21% 16/75 [00:04<00:16,  3.67it/s]
 23% 17/75 [00:04<00:14,  3.92it/s]
 24% 18/75 [00:04<00:14,  3.97it/s]
 25% 19/75 [00:04<00:15,  3.72it/s]
 27% 20/75 [00:05<00:15,  3.65it/s]
 28% 21/75 [00:05<00:14,  3.63it/s]
 29% 22/75 [00:05<00:15,  3.48it/s]
 31% 23/75 [00:06<00:15,  3.36it/s]
 32% 24/75 [00:06<00:14,  3.51it/s]
 33% 25/75 [00:06<00:13,  3.74it/s]
 35% 26/75 [00:06<00:12,  3.88it/s]
 36% 27/75 [00:07<00:12,  3.78it/s]
 37% 28/75 [00:07<00:12,  3.70it/s]
 39% 29/75 [00:07<00:12,  3.57it/s]
 40% 30/75 [00:08<00:13,  3.38it/s]
 41% 31/75 [00:08<00:13,  3.25it/s]
 43% 32/75 [00:08<00:11,  3.67it/s]
 44% 33/75 [00:08<00:11,  3.58it/s]
 45% 34/75 [00:09<00:11,  3.52it/s]
 47% 35/75 [00:09<00:11,  3.39it/s]
 48% 36/75 [00:09<00:11,  3.34it/s]
 49% 37/75 [00:10<00:11,  3.43it/s]
 51% 38/75 [00:10<00:10,  3.45it/s]
 52% 39/75 [00:10<00:10,  3.49it/s]
 53% 40/75 [00:10<00:10,  3.44it/s]
 55% 41/75 [00:11<00:09,  3.52it/s]
 56% 42/75 [00:11<00:09,  3.49it/s]
 57% 43/75 [00:11<00:09,  3.53it/s]
 59% 44/75 [00:12<00:08,  3.77it/s]
 60% 45/75 [00:12<00:08,  3.72it/s]
 61% 46/75 [00:12<00:08,  3.42it/s]
 63% 47/75 [00:13<00:08,  3.25it/s]
 64% 48/75 [00:13<00:08,  3.28it/s]
 65% 49/75 [00:13<00:07,  3.28it/s]
 67% 50/75 [00:13<00:07,  3.25it/s]
 68% 51/75 [00:14<00:07,  3.39it/s]
 69% 52/75 [00:14<00:06,  3.53it/s]
 71% 53/75 [00:14<00:07,  3.13it/s]
 72% 54/75 [00:15<00:06,  3.16it/s]
 73% 55/75 [00:15<00:05,  3.49it/s]
 75% 56/75 [00:15<00:05,  3.44it/s]
 76% 57/75 [00:15<00:05,  3.40it/s]
 77% 58/75 [00:16<00:04,  3.54it/s]
 79% 59/75 [00:16<00:04,  3.70it/s]
 80% 60/75 [00:16<00:04,  3.54it/s]
 81% 61/75 [00:17<00:04,  3.45it/s]
 83% 62/75 [00:17<00:03,  3.35it/s]
 84% 63/75 [00:17<00:03,  3.39it/s]
 85% 64/75 [00:17<00:03,  3.45it/s]
 87% 65/75 [00:18<00:02,  3.62it/s]
 88% 66/75 [00:18<00:02,  3.95it/s]
 89% 67/75 [00:18<00:01,  4.06it/s]
 91% 68/75 [00:18<00:01,  3.87it/s]
 92% 69/75 [00:19<00:01,  3.86it/s]
 93% 70/75 [00:19<00:01,  4.11it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.92it/s]
 97% 73/75 [00:20<00:00,  3.99it/s]
 99% 74/75 [00:20<00:00,  3.76it/s]
                                         
{'eval_loss': 1.3309905529022217, 'eval_runtime': 21.009, 'eval_samples_per_second': 3.57, 'eval_steps_per_second': 3.57, 'epoch': 0.01}
  0% 70/22194 [06:12<17:23:21,  2.83s/it]
100% 75/75 [00:20<00:00,  3.70it/s]
{'loss': 1.4298, 'learning_rate': 0.00019999626881308364, 'epoch': 0.01}
{'loss': 1.4024, 'learning_rate': 0.00019999614547714108, 'epoch': 0.01}
{'loss': 1.6378, 'learning_rate': 0.00019999602013578792, 'epoch': 0.01}
{'loss': 1.8874, 'learning_rate': 0.0001999958927890268, 'epoch': 0.01}
{'loss': 1.1798, 'learning_rate': 0.00019999576343686018, 'epoch': 0.01}
  0% 75/22194 [06:24<19:48:31,  3.22s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 11.32it/s]
  5% 4/75 [00:00<00:13,  5.39it/s]
  7% 5/75 [00:00<00:14,  4.71it/s]
  8% 6/75 [00:01<00:14,  4.64it/s]
  9% 7/75 [00:01<00:16,  4.02it/s]
 11% 8/75 [00:01<00:17,  3.89it/s]
 12% 9/75 [00:02<00:16,  4.10it/s]
 13% 10/75 [00:02<00:15,  4.09it/s]
 15% 11/75 [00:02<00:15,  4.24it/s]
 16% 12/75 [00:02<00:15,  4.10it/s]
 17% 13/75 [00:02<00:15,  4.02it/s]
 19% 14/75 [00:03<00:14,  4.09it/s]
 20% 15/75 [00:03<00:14,  4.15it/s]
 21% 16/75 [00:03<00:15,  3.87it/s]
 23% 17/75 [00:03<00:14,  3.97it/s]
 24% 18/75 [00:04<00:13,  4.15it/s]
 25% 19/75 [00:04<00:14,  3.94it/s]
 27% 20/75 [00:04<00:14,  3.80it/s]
 28% 21/75 [00:05<00:14,  3.78it/s]
 29% 22/75 [00:05<00:14,  3.74it/s]
 31% 23/75 [00:05<00:14,  3.52it/s]
 32% 24/75 [00:05<00:13,  3.66it/s]
 33% 25/75 [00:06<00:12,  3.93it/s]
 35% 26/75 [00:06<00:11,  4.15it/s]
 36% 27/75 [00:06<00:11,  4.09it/s]
 37% 28/75 [00:06<00:12,  3.88it/s]
 39% 29/75 [00:07<00:11,  3.87it/s]
 40% 30/75 [00:07<00:12,  3.64it/s]
 41% 31/75 [00:07<00:12,  3.42it/s]
 43% 32/75 [00:07<00:11,  3.82it/s]
 44% 33/75 [00:08<00:10,  3.86it/s]
 45% 34/75 [00:08<00:11,  3.68it/s]
 47% 35/75 [00:08<00:11,  3.54it/s]
 48% 36/75 [00:09<00:10,  3.55it/s]
 49% 37/75 [00:09<00:10,  3.61it/s]
 51% 38/75 [00:09<00:10,  3.70it/s]
 52% 39/75 [00:09<00:09,  3.62it/s]
 53% 40/75 [00:10<00:09,  3.63it/s]
 55% 41/75 [00:10<00:09,  3.71it/s]
 56% 42/75 [00:10<00:09,  3.65it/s]
 57% 43/75 [00:10<00:08,  3.70it/s]
 59% 44/75 [00:11<00:07,  4.05it/s]
 60% 45/75 [00:11<00:07,  3.85it/s]
 61% 46/75 [00:11<00:08,  3.56it/s]
 63% 47/75 [00:12<00:08,  3.37it/s]
 64% 48/75 [00:12<00:08,  3.36it/s]
 65% 49/75 [00:12<00:07,  3.42it/s]
 67% 50/75 [00:13<00:07,  3.37it/s]
 68% 51/75 [00:13<00:06,  3.52it/s]
 69% 52/75 [00:13<00:06,  3.61it/s]
 71% 53/75 [00:13<00:06,  3.20it/s]
 72% 54/75 [00:14<00:06,  3.25it/s]
 73% 55/75 [00:14<00:05,  3.55it/s]
 75% 56/75 [00:14<00:05,  3.49it/s]
 76% 57/75 [00:15<00:05,  3.44it/s]
 77% 58/75 [00:15<00:04,  3.61it/s]
 79% 59/75 [00:15<00:04,  3.80it/s]
 80% 60/75 [00:15<00:04,  3.58it/s]
 81% 61/75 [00:16<00:04,  3.49it/s]
 83% 62/75 [00:16<00:03,  3.39it/s]
 84% 63/75 [00:16<00:03,  3.43it/s]
 85% 64/75 [00:17<00:03,  3.48it/s]
 87% 65/75 [00:17<00:02,  3.60it/s]
 88% 66/75 [00:17<00:02,  3.95it/s]
 89% 67/75 [00:17<00:01,  4.07it/s]
 91% 68/75 [00:17<00:01,  3.86it/s]
 92% 69/75 [00:18<00:01,  3.87it/s]
 93% 70/75 [00:18<00:01,  4.05it/s]
 95% 71/75 [00:18<00:01,  3.92it/s]
 96% 72/75 [00:18<00:00,  3.88it/s]
 97% 73/75 [00:19<00:00,  3.94it/s]
 99% 74/75 [00:19<00:00,  3.71it/s]
                                         
{'eval_loss': 1.3255985975265503, 'eval_runtime': 20.0817, 'eval_samples_per_second': 3.735, 'eval_steps_per_second': 3.735, 'epoch': 0.01}
  0% 75/22194 [06:44<19:48:31,  3.22s/it]
100% 75/75 [00:20<00:00,  3.65it/s]
{'loss': 0.9032, 'learning_rate': 0.0001999956320792907, 'epoch': 0.01}
{'loss': 1.7063, 'learning_rate': 0.00019999549871632102, 'epoch': 0.01}
{'loss': 1.3649, 'learning_rate': 0.00019999536334795373, 'epoch': 0.01}
{'loss': 1.0759, 'learning_rate': 0.0001999952259741916, 'epoch': 0.01}
{'loss': 1.5272, 'learning_rate': 0.00019999508659503741, 'epoch': 0.01}
  0% 80/22194 [06:48<16:46:26,  2.73s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.08it/s]
  5% 4/75 [00:00<00:12,  5.66it/s]
  7% 5/75 [00:01<00:16,  4.30it/s]
  8% 6/75 [00:01<00:15,  4.55it/s]
  9% 7/75 [00:01<00:17,  3.96it/s]
 11% 8/75 [00:01<00:18,  3.59it/s]
 12% 9/75 [00:02<00:17,  3.85it/s]
 13% 10/75 [00:02<00:17,  3.80it/s]
 15% 11/75 [00:02<00:15,  4.11it/s]
 16% 12/75 [00:02<00:16,  3.93it/s]
 17% 13/75 [00:03<00:16,  3.85it/s]
 19% 14/75 [00:03<00:15,  3.90it/s]
 20% 15/75 [00:03<00:15,  3.97it/s]
 21% 16/75 [00:03<00:16,  3.65it/s]
 23% 17/75 [00:04<00:15,  3.74it/s]
 24% 18/75 [00:04<00:14,  3.92it/s]
 25% 19/75 [00:04<00:15,  3.72it/s]
 27% 20/75 [00:05<00:15,  3.63it/s]
 28% 21/75 [00:05<00:14,  3.62it/s]
 29% 22/75 [00:05<00:14,  3.55it/s]
 31% 23/75 [00:05<00:15,  3.38it/s]
 32% 24/75 [00:06<00:14,  3.50it/s]
 33% 25/75 [00:06<00:13,  3.83it/s]
 35% 26/75 [00:06<00:12,  3.97it/s]
 36% 27/75 [00:06<00:12,  3.81it/s]
 37% 28/75 [00:07<00:12,  3.70it/s]
 39% 29/75 [00:07<00:12,  3.64it/s]
 40% 30/75 [00:07<00:13,  3.42it/s]
 41% 31/75 [00:08<00:12,  3.40it/s]
 43% 32/75 [00:08<00:11,  3.68it/s]
 44% 33/75 [00:08<00:11,  3.64it/s]
 45% 34/75 [00:08<00:11,  3.52it/s]
 47% 35/75 [00:09<00:11,  3.43it/s]
 48% 36/75 [00:09<00:11,  3.46it/s]
 49% 37/75 [00:09<00:10,  3.46it/s]
 51% 38/75 [00:10<00:10,  3.47it/s]
 52% 39/75 [00:10<00:10,  3.49it/s]
 53% 40/75 [00:10<00:09,  3.52it/s]
 55% 41/75 [00:10<00:09,  3.51it/s]
 56% 42/75 [00:11<00:09,  3.51it/s]
 57% 43/75 [00:11<00:08,  3.61it/s]
 59% 44/75 [00:11<00:08,  3.79it/s]
 60% 45/75 [00:11<00:08,  3.65it/s]
 61% 46/75 [00:12<00:08,  3.40it/s]
 63% 47/75 [00:12<00:08,  3.25it/s]
 64% 48/75 [00:12<00:08,  3.23it/s]
 65% 49/75 [00:13<00:07,  3.29it/s]
 67% 50/75 [00:13<00:07,  3.25it/s]
 68% 51/75 [00:13<00:07,  3.36it/s]
 69% 52/75 [00:14<00:06,  3.55it/s]
 71% 53/75 [00:14<00:07,  3.14it/s]
 72% 54/75 [00:14<00:06,  3.08it/s]
 73% 55/75 [00:15<00:05,  3.50it/s]1

 75% 56/75 [00:15<00:05,  3.43it/s]
 76% 57/75 [00:15<00:05,  3.36it/s]
 77% 58/75 [00:15<00:04,  3.41it/s]
 79% 59/75 [00:16<00:04,  3.78it/s]
 80% 60/75 [00:16<00:04,  3.53it/s]1

 81% 61/75 [00:16<00:04,  3.49it/s]
 83% 62/75 [00:17<00:03,  3.32it/s]1

 84% 63/75 [00:17<00:03,  3.35it/s]
 85% 64/75 [00:17<00:03,  3.47it/s]
 87% 65/75 [00:17<00:02,  3.51it/s]
 88% 66/75 [00:18<00:02,  3.84it/s]
 89% 67/75 [00:18<00:01,  4.12it/s]
 91% 68/75 [00:18<00:01,  3.80it/s]
 92% 69/75 [00:18<00:01,  3.77it/s]
 93% 70/75 [00:19<00:01,  4.14it/s]
 95% 71/75 [00:19<00:00,  4.04it/s]
 96% 72/75 [00:19<00:00,  3.98it/s]
 97% 73/75 [00:19<00:00,  3.98it/s]
 99% 74/75 [00:20<00:00,  3.73it/s]
                                         
{'eval_loss': 1.3223317861557007, 'eval_runtime': 20.7687, 'eval_samples_per_second': 3.611, 'eval_steps_per_second': 3.611, 'epoch': 0.01}
  0% 80/22194 [07:09<16:46:26,  2.73s/it]
100% 75/75 [00:20<00:00,  3.67it/s]
                                   1
2
3
y









{'loss': 1.1238, 'learning_rate': 0.0001999949452104939, 'epoch': 0.01}
{'loss': 1.2245, 'learning_rate': 0.00019999480182056396, 'epoch': 0.01}
{'loss': 1.5251, 'learning_rate': 0.0001999946564252504, 'epoch': 0.01}
{'loss': 1.3594, 'learning_rate': 0.0001999945090245562, 'epoch': 0.01}
{'loss': 0.4523, 'learning_rate': 0.00019999435961848424, 'epoch': 0.01}
  0% 85/22194 [07:22<20:02:04,  3.26s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.89it/s]
  4% 3/75 [00:00<00:09,  7.31it/s]
  5% 4/75 [00:00<00:13,  5.44it/s]
  7% 5/75 [00:00<00:16,  4.29it/s]
  8% 6/75 [00:01<00:15,  4.53it/s]
  9% 7/75 [00:01<00:17,  3.95it/s]
 11% 8/75 [00:01<00:18,  3.66it/s]
 12% 9/75 [00:02<00:16,  3.99it/s]
 13% 10/75 [00:02<00:16,  4.03it/s]
 15% 11/75 [00:02<00:15,  4.24it/s]
 16% 12/75 [00:02<00:15,  4.06it/s]
 17% 13/75 [00:03<00:15,  3.90it/s]
 19% 14/75 [00:03<00:15,  3.99it/s]
 20% 15/75 [00:03<00:14,  4.10it/s]
 21% 16/75 [00:03<00:15,  3.80it/s]
 23% 17/75 [00:04<00:15,  3.84it/s]
 24% 18/75 [00:04<00:14,  4.01it/s]
 25% 19/75 [00:04<00:14,  3.88it/s]
 27% 20/75 [00:04<00:14,  3.75it/s]
 28% 21/75 [00:05<00:14,  3.73it/s]
 29% 22/75 [00:05<00:14,  3.66it/s]
 31% 23/75 [00:05<00:14,  3.53it/s]
 32% 24/75 [00:05<00:14,  3.63it/s]
 33% 25/75 [00:06<00:12,  3.94it/s]
 35% 26/75 [00:06<00:11,  4.12it/s]
 36% 27/75 [00:06<00:12,  3.97it/s]
 37% 28/75 [00:06<00:12,  3.80it/s]
 39% 29/75 [00:07<00:12,  3.72it/s]
 40% 30/75 [00:07<00:12,  3.53it/s]
 41% 31/75 [00:07<00:12,  3.52it/s]
 43% 32/75 [00:08<00:11,  3.74it/s]
 44% 33/75 [00:08<00:11,  3.74it/s]
 45% 34/75 [00:08<00:11,  3.61it/s]
 47% 35/75 [00:08<00:11,  3.50it/s]
 48% 36/75 [00:09<00:11,  3.51it/s]
 49% 37/75 [00:09<00:10,  3.54it/s]
 51% 38/75 [00:09<00:10,  3.58it/s]
 52% 39/75 [00:10<00:10,  3.55it/s]
 53% 40/75 [00:10<00:09,  3.60it/s]
 55% 41/75 [00:10<00:09,  3.60it/s]
 56% 42/75 [00:10<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.66it/s]
 59% 44/75 [00:11<00:08,  3.85it/s]
 60% 45/75 [00:11<00:08,  3.68it/s]
 61% 46/75 [00:11<00:08,  3.44it/s]
 63% 47/75 [00:12<00:08,  3.27it/s]
 64% 48/75 [00:12<00:08,  3.23it/s]
 65% 49/75 [00:12<00:07,  3.37it/s]
 67% 50/75 [00:13<00:07,  3.31it/s]
 68% 51/75 [00:13<00:07,  3.33it/s]
 69% 52/75 [00:13<00:06,  3.57it/s]
 71% 53/75 [00:14<00:06,  3.19it/s]
 72% 54/75 [00:14<00:06,  3.17it/s]
 73% 55/75 [00:14<00:05,  3.47it/s]
 75% 56/75 [00:14<00:05,  3.44it/s]
 76% 57/75 [00:15<00:05,  3.32it/s]
 77% 58/75 [00:15<00:04,  3.46it/s]
 79% 59/75 [00:15<00:04,  3.80it/s]
 80% 60/75 [00:16<00:04,  3.53it/s]
 81% 61/75 [00:16<00:04,  3.48it/s]
 83% 62/75 [00:16<00:03,  3.38it/s]
 84% 63/75 [00:17<00:03,  3.40it/s]
 85% 64/75 [00:17<00:03,  3.44it/s]
 87% 65/75 [00:17<00:02,  3.50it/s]
 88% 66/75 [00:17<00:02,  3.88it/s]
 89% 67/75 [00:17<00:01,  4.08it/s]
 91% 68/75 [00:18<00:01,  3.95it/s]
 92% 69/75 [00:18<00:01,  3.93it/s]
 93% 70/75 [00:18<00:01,  4.08it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.90it/s]
 97% 73/75 [00:19<00:00,  3.93it/s]
 99% 74/75 [00:19<00:00,  3.70it/s]
                                         
{'eval_loss': 1.3399217128753662, 'eval_runtime': 20.4168, 'eval_samples_per_second': 3.673, 'eval_steps_per_second': 3.673, 'epoch': 0.01}
  0% 85/22194 [07:42<20:02:04,  3.26s/it]
100% 75/75 [00:20<00:00,  3.65it/s]
{'loss': 1.4205, 'learning_rate': 0.00019999420820703763, 'epoch': 0.01}
{'loss': 1.0825, 'learning_rate': 0.00019999405479021931, 'epoch': 0.01}
{'loss': 1.1175, 'learning_rate': 0.00019999389936803237, 'epoch': 0.01}
{'loss': 1.1628, 'learning_rate': 0.00019999374194047998, 'epoch': 0.01}
{'loss': 0.812, 'learning_rate': 0.00019999358250756525, 'epoch': 0.01}
  0% 90/22194 [07:46<16:43:10,  2.72s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.38it/s]
  5% 4/75 [00:00<00:13,  5.19it/s]
  7% 5/75 [00:01<00:16,  4.37it/s]
  8% 6/75 [00:01<00:15,  4.39it/s]
  9% 7/75 [00:01<00:17,  3.83it/s]
 11% 8/75 [00:01<00:18,  3.65it/s]
 12% 9/75 [00:02<00:17,  3.87it/s]
 13% 10/75 [00:02<00:16,  3.90it/s]
 15% 11/75 [00:02<00:15,  4.05it/s]
 16% 12/75 [00:02<00:16,  3.89it/s]
 17% 13/75 [00:03<00:16,  3.80it/s]
 19% 14/75 [00:03<00:15,  3.84it/s]
 20% 15/75 [00:03<00:15,  3.93it/s]
 21% 16/75 [00:03<00:16,  3.64it/s]
 23% 17/75 [00:04<00:15,  3.66it/s]
 24% 18/75 [00:04<00:14,  4.02it/s]
 25% 19/75 [00:04<00:14,  3.81it/s]
 27% 20/75 [00:05<00:15,  3.67it/s]
 28% 21/75 [00:05<00:14,  3.61it/s]
 29% 22/75 [00:05<00:15,  3.48it/s]
 31% 23/75 [00:05<00:15,  3.44it/s]
 32% 24/75 [00:06<00:14,  3.53it/s]
 33% 25/75 [00:06<00:13,  3.79it/s]
 35% 26/75 [00:06<00:12,  4.06it/s]
 36% 27/75 [00:06<00:12,  3.87it/s]
 37% 28/75 [00:07<00:12,  3.73it/s]
 39% 29/75 [00:07<00:12,  3.69it/s]
 40% 30/75 [00:07<00:13,  3.41it/s]
 41% 31/75 [00:08<00:13,  3.36it/s]
 43% 32/75 [00:08<00:11,  3.69it/s]
 44% 33/75 [00:08<00:11,  3.65it/s]
 45% 34/75 [00:08<00:11,  3.54it/s]
 47% 35/75 [00:09<00:11,  3.43it/s]
 48% 36/75 [00:09<00:11,  3.48it/s]
 49% 37/75 [00:09<00:10,  3.48it/s]
 51% 38/75 [00:10<00:10,  3.50it/s]
 52% 39/75 [00:10<00:10,  3.52it/s]
 53% 40/75 [00:10<00:09,  3.54it/s]
 55% 41/75 [00:10<00:09,  3.56it/s]
 56% 42/75 [00:11<00:09,  3.53it/s]
 57% 43/75 [00:11<00:08,  3.63it/s]
 59% 44/75 [00:11<00:08,  3.82it/s]
 60% 45/75 [00:11<00:08,  3.68it/s]
 61% 46/75 [00:12<00:08,  3.42it/s]
 63% 47/75 [00:12<00:08,  3.26it/s]
 64% 48/75 [00:12<00:08,  3.26it/s]
 65% 49/75 [00:13<00:07,  3.29it/s]
 67% 50/75 [00:13<00:07,  3.27it/s]
 68% 51/75 [00:13<00:06,  3.44it/s]
 69% 52/75 [00:14<00:06,  3.52it/s]
 71% 53/75 [00:14<00:07,  3.13it/s]
 72% 54/75 [00:14<00:06,  3.18it/s]
 73% 55/75 [00:15<00:05,  3.44it/s]
 75% 56/75 [00:15<00:05,  3.40it/s]
 76% 57/75 [00:15<00:05,  3.37it/s]
 77% 58/75 [00:15<00:04,  3.53it/s]
 79% 59/75 [00:16<00:04,  3.67it/s]
 80% 60/75 [00:16<00:04,  3.52it/s]
 81% 61/75 [00:16<00:04,  3.45it/s]
 83% 62/75 [00:17<00:03,  3.33it/s]
 84% 63/75 [00:17<00:03,  3.37it/s]
 85% 64/75 [00:17<00:03,  3.45it/s]
 87% 65/75 [00:17<00:02,  3.57it/s]
 88% 66/75 [00:18<00:02,  3.84it/s]
 89% 67/75 [00:18<00:01,  4.02it/s]
 91% 68/75 [00:18<00:01,  3.85it/s]
 92% 69/75 [00:18<00:01,  3.87it/s]
 93% 70/75 [00:19<00:01,  4.07it/s]
 95% 71/75 [00:19<00:01,  3.94it/s]
 96% 72/75 [00:19<00:00,  3.91it/s]
 97% 73/75 [00:19<00:00,  3.93it/s]
 99% 74/75 [00:20<00:00,  3.66it/s]
                                         
{'eval_loss': 1.3145769834518433, 'eval_runtime': 20.7113, 'eval_samples_per_second': 3.621, 'eval_steps_per_second': 3.621, 'epoch': 0.01}
  0% 90/22194 [08:07<16:43:10,  2.72s/it]
100% 75/75 [00:20<00:00,  3.68it/s]
{'loss': 0.7807, 'learning_rate': 0.00019999342106929137, 'epoch': 0.01}
{'loss': 1.4437, 'learning_rate': 0.00019999325762566162, 'epoch': 0.01}
{'loss': 1.491, 'learning_rate': 0.00019999309217667922, 'epoch': 0.01}
{'loss': 1.123, 'learning_rate': 0.00019999292472234754, 'epoch': 0.01}
{'loss': 1.4219, 'learning_rate': 0.00019999275526266994, 'epoch': 0.01}
  0% 95/22194 [08:23<21:44:53,  3.54s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.56it/s]
  5% 4/75 [00:00<00:12,  5.79it/s]
  7% 5/75 [00:00<00:15,  4.49it/s]
  8% 6/75 [00:01<00:14,  4.69it/s]
  9% 7/75 [00:01<00:16,  4.06it/s]
 11% 8/75 [00:01<00:17,  3.80it/s]
 12% 9/75 [00:02<00:16,  4.00it/s]
 13% 10/75 [00:02<00:16,  3.97it/s]
 15% 11/75 [00:02<00:15,  4.23it/s]
 16% 12/75 [00:02<00:15,  4.13it/s]
 17% 13/75 [00:03<00:15,  3.95it/s]
 19% 14/75 [00:03<00:15,  3.91it/s]
 20% 15/75 [00:03<00:14,  4.23it/s]
 21% 16/75 [00:03<00:15,  3.71it/s]
 23% 17/75 [00:04<00:15,  3.79it/s]
 24% 18/75 [00:04<00:13,  4.31it/s]
 25% 19/75 [00:04<00:13,  4.07it/s]
 27% 20/75 [00:04<00:14,  3.87it/s]
 28% 21/75 [00:05<00:14,  3.80it/s]
 29% 22/75 [00:05<00:14,  3.73it/s]
 31% 23/75 [00:05<00:14,  3.57it/s]
 32% 24/75 [00:05<00:13,  3.66it/s]
 33% 25/75 [00:06<00:12,  3.96it/s]
 35% 26/75 [00:06<00:11,  4.14it/s]
 36% 27/75 [00:06<00:12,  3.97it/s]
 37% 28/75 [00:06<00:12,  3.83it/s]
 39% 29/75 [00:07<00:12,  3.74it/s]
 40% 30/75 [00:07<00:12,  3.54it/s]
 41% 31/75 [00:07<00:12,  3.53it/s]
 43% 32/75 [00:07<00:11,  3.75it/s]
 44% 33/75 [00:08<00:11,  3.77it/s]
 45% 34/75 [00:08<00:11,  3.62it/s]
 47% 35/75 [00:08<00:11,  3.52it/s]
 48% 36/75 [00:09<00:11,  3.53it/s]
 49% 37/75 [00:09<00:10,  3.53it/s]
 51% 38/75 [00:09<00:10,  3.54it/s]
 52% 39/75 [00:10<00:10,  3.51it/s]
 53% 40/75 [00:10<00:09,  3.56it/s]
 55% 41/75 [00:10<00:09,  3.56it/s]
 56% 42/75 [00:10<00:09,  3.53it/s]
 57% 43/75 [00:11<00:08,  3.64it/s]
 59% 44/75 [00:11<00:08,  3.83it/s]
 60% 45/75 [00:11<00:08,  3.66it/s]
 61% 46/75 [00:11<00:08,  3.41it/s]
 63% 47/75 [00:12<00:08,  3.24it/s]
 64% 48/75 [00:12<00:08,  3.24it/s]
 65% 49/75 [00:12<00:07,  3.26it/s]
 67% 50/75 [00:13<00:07,  3.24it/s]
 68% 51/75 [00:13<00:07,  3.39it/s]
 69% 52/75 [00:13<00:06,  3.48it/s]
 71% 53/75 [00:14<00:07,  3.09it/s]
 72% 54/75 [00:14<00:06,  3.05it/s]
 73% 55/75 [00:14<00:05,  3.43it/s]
 75% 56/75 [00:15<00:05,  3.38it/s]
 76% 57/75 [00:15<00:05,  3.30it/s]
 77% 58/75 [00:15<00:05,  3.37it/s]
 79% 59/75 [00:15<00:04,  3.63it/s]
 80% 60/75 [00:16<00:04,  3.47it/s]
 81% 61/75 [00:16<00:04,  3.38it/s]
 83% 62/75 [00:16<00:04,  3.24it/s]
 84% 63/75 [00:17<00:03,  3.25it/s]
 85% 64/75 [00:17<00:03,  3.34it/s]
 87% 65/75 [00:17<00:02,  3.42it/s]
 88% 66/75 [00:17<00:02,  3.73it/s]
 89% 67/75 [00:18<00:02,  3.98it/s]
 91% 68/75 [00:18<00:01,  3.74it/s]
 92% 69/75 [00:18<00:01,  3.66it/s]
 93% 70/75 [00:18<00:01,  3.89it/s]
 95% 71/75 [00:19<00:01,  3.89it/s]
 96% 72/75 [00:19<00:00,  3.65it/s]
 97% 73/75 [00:19<00:00,  3.87it/s]
 99% 74/75 [00:20<00:00,  3.55it/s]
                                         
{'eval_loss': 1.296973705291748, 'eval_runtime': 20.6143, 'eval_samples_per_second': 3.638, 'eval_steps_per_second': 3.638, 'epoch': 0.01}
  0% 95/22194 [08:44<21:44:53,  3.54s/it]
100% 75/75 [00:20<00:00,  3.56it/s]
{'loss': 1.7721, 'learning_rate': 0.00019999258379764978, 'epoch': 0.01}
{'loss': 0.9856, 'learning_rate': 0.00019999241032729056, 'epoch': 0.01}
{'loss': 1.4707, 'learning_rate': 0.00019999223485159566, 'epoch': 0.01}
{'loss': 0.8617, 'learning_rate': 0.00019999205737056868, 'epoch': 0.01}
{'loss': 0.9788, 'learning_rate': 0.00019999187788421314, 'epoch': 0.01}
  0% 100/22194 [08:49<17:27:14,  2.84s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.71it/s]
  4% 3/75 [00:00<00:11,  6.28it/s]
  5% 4/75 [00:00<00:15,  4.72it/s]
  7% 5/75 [00:01<00:17,  3.96it/s]
  8% 6/75 [00:01<00:16,  4.08it/s]
  9% 7/75 [00:01<00:18,  3.62it/s]
 11% 8/75 [00:01<00:19,  3.51it/s]
 12% 9/75 [00:02<00:17,  3.72it/s]
 13% 10/75 [00:02<00:17,  3.76it/s]
 15% 11/75 [00:02<00:16,  3.94it/s]
 16% 12/75 [00:02<00:16,  3.79it/s]
 17% 13/75 [00:03<00:16,  3.72it/s]
 19% 14/75 [00:03<00:16,  3.78it/s]
 20% 15/75 [00:03<00:15,  3.82it/s]
 21% 16/75 [00:04<00:16,  3.59it/s]
 23% 17/75 [00:04<00:15,  3.69it/s]
 24% 18/75 [00:04<00:14,  3.88it/s]
 25% 19/75 [00:04<00:15,  3.70it/s]
 27% 20/75 [00:05<00:15,  3.60it/s]
 28% 21/75 [00:05<00:15,  3.57it/s]
 29% 22/75 [00:05<00:14,  3.54it/s]
 31% 23/75 [00:06<00:15,  3.40it/s]
 32% 24/75 [00:06<00:14,  3.46it/s]
 33% 25/75 [00:06<00:13,  3.78it/s]
 35% 26/75 [00:06<00:12,  4.02it/s]
 36% 27/75 [00:07<00:12,  3.79it/s]
 37% 28/75 [00:07<00:12,  3.73it/s]
 39% 29/75 [00:07<00:12,  3.65it/s]
 40% 30/75 [00:07<00:13,  3.44it/s]
 41% 31/75 [00:08<00:13,  3.30it/s]
 43% 32/75 [00:08<00:11,  3.67it/s]
 44% 33/75 [00:08<00:11,  3.66it/s]
 45% 34/75 [00:09<00:11,  3.53it/s]
 47% 35/75 [00:09<00:11,  3.42it/s]
 48% 36/75 [00:09<00:11,  3.46it/s]
 49% 37/75 [00:09<00:11,  3.38it/s]
 51% 38/75 [00:10<00:10,  3.44it/s]
 52% 39/75 [00:10<00:10,  3.50it/s]
 53% 40/75 [00:10<00:09,  3.56it/s]
 55% 41/75 [00:11<00:09,  3.59it/s]
 56% 42/75 [00:11<00:09,  3.53it/s]
 57% 43/75 [00:11<00:08,  3.66it/s]
 59% 44/75 [00:11<00:08,  3.85it/s]
 60% 45/75 [00:12<00:08,  3.68it/s]
 61% 46/75 [00:12<00:08,  3.42it/s]
 63% 47/75 [00:12<00:08,  3.27it/s]
 64% 48/75 [00:13<00:08,  3.26it/s]
 65% 49/75 [00:13<00:07,  3.29it/s]
 67% 50/75 [00:13<00:07,  3.27it/s]
 68% 51/75 [00:13<00:06,  3.44it/s]
 69% 52/75 [00:14<00:06,  3.52it/s]
 71% 53/75 [00:14<00:07,  3.13it/s]
 72% 54/75 [00:14<00:06,  3.18it/s]
 73% 55/75 [00:15<00:05,  3.46it/s]
 75% 56/75 [00:15<00:05,  3.39it/s]
 76% 57/75 [00:15<00:05,  3.36it/s]
 77% 58/75 [00:16<00:04,  3.52it/s]
 79% 59/75 [00:16<00:04,  3.71it/s]
 80% 60/75 [00:16<00:04,  3.56it/s]
 81% 61/75 [00:16<00:04,  3.47it/s]
 83% 62/75 [00:17<00:03,  3.36it/s]
 84% 63/75 [00:17<00:03,  3.42it/s]
 85% 64/75 [00:17<00:03,  3.49it/s]
 87% 65/75 [00:18<00:02,  3.61it/s]
 88% 66/75 [00:18<00:02,  3.94it/s]
 89% 67/75 [00:18<00:01,  4.04it/s]
 91% 68/75 [00:18<00:01,  3.86it/s]
 92% 69/75 [00:18<00:01,  3.89it/s]
 93% 70/75 [00:19<00:01,  4.08it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.89it/s]
 97% 73/75 [00:19<00:00,  3.99it/s]
 99% 74/75 [00:20<00:00,  3.76it/s]
                                          
{'eval_loss': 1.311365008354187, 'eval_runtime': 20.7956, 'eval_samples_per_second': 3.607, 'eval_steps_per_second': 3.607, 'epoch': 0.01}
  0% 100/22194 [09:09<17:27:14,  2.84s/it]
100% 75/75 [00:20<00:00,  3.70it/s]
{'loss': 1.3766, 'learning_rate': 0.00019999169639253265, 'epoch': 0.01}
{'loss': 1.313, 'learning_rate': 0.00019999151289553087, 'epoch': 0.01}
{'loss': 0.9775, 'learning_rate': 0.0001999913273932115, 'epoch': 0.01}
{'loss': 1.644, 'learning_rate': 0.0001999911398855782, 'epoch': 0.01}
{'loss': 1.0716, 'learning_rate': 0.00019999095037263474, 'epoch': 0.01}
  0% 105/22194 [09:21<19:43:58,  3.22s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.21it/s]
  4% 3/75 [00:00<00:09,  7.47it/s]
  5% 4/75 [00:00<00:13,  5.42it/s]
  7% 5/75 [00:00<00:16,  4.34it/s]
  8% 6/75 [00:01<00:15,  4.53it/s]
  9% 7/75 [00:01<00:17,  3.98it/s]
 11% 8/75 [00:01<00:18,  3.65it/s]
 12% 9/75 [00:01<00:16,  4.05it/s]
 13% 10/75 [00:02<00:16,  4.00it/s]
 15% 11/75 [00:02<00:15,  4.27it/s]
 16% 12/75 [00:02<00:15,  4.17it/s]
 17% 13/75 [00:02<00:15,  4.04it/s]
 19% 14/75 [00:03<00:15,  3.96it/s]
 20% 15/75 [00:03<00:13,  4.29it/s]
 21% 16/75 [00:03<00:15,  3.80it/s]
 23% 17/75 [00:03<00:14,  3.90it/s]
 24% 18/75 [00:04<00:13,  4.15it/s]
 25% 19/75 [00:04<00:14,  3.96it/s]
 27% 20/75 [00:04<00:14,  3.80it/s]
 28% 21/75 [00:05<00:14,  3.72it/s]
 29% 22/75 [00:05<00:13,  3.87it/s]
 31% 23/75 [00:05<00:14,  3.64it/s]
 32% 24/75 [00:05<00:13,  3.68it/s]
 33% 25/75 [00:06<00:12,  3.97it/s]
 35% 26/75 [00:06<00:11,  4.22it/s]
 36% 27/75 [00:06<00:11,  4.14it/s]
 37% 28/75 [00:06<00:11,  3.96it/s]
 39% 29/75 [00:07<00:11,  3.84it/s]
 40% 30/75 [00:07<00:12,  3.59it/s]
 41% 31/75 [00:07<00:12,  3.52it/s]
 43% 32/75 [00:07<00:11,  3.81it/s]
 44% 33/75 [00:08<00:11,  3.80it/s]
 45% 34/75 [00:08<00:11,  3.71it/s]
 47% 35/75 [00:08<00:11,  3.57it/s]
 48% 36/75 [00:09<00:10,  3.58it/s]
 49% 37/75 [00:09<00:10,  3.62it/s]
 51% 38/75 [00:09<00:10,  3.62it/s]
 52% 39/75 [00:09<00:09,  3.64it/s]
 53% 40/75 [00:10<00:09,  3.66it/s]
 55% 41/75 [00:10<00:09,  3.60it/s]
 56% 42/75 [00:10<00:08,  3.74it/s]
 57% 43/75 [00:10<00:08,  3.82it/s]
 59% 44/75 [00:11<00:07,  4.01it/s]
 60% 45/75 [00:11<00:07,  3.81it/s]
 61% 46/75 [00:11<00:08,  3.53it/s]
 63% 47/75 [00:12<00:08,  3.35it/s]
 64% 48/75 [00:12<00:08,  3.34it/s]
 65% 49/75 [00:12<00:07,  3.37it/s]
 67% 50/75 [00:12<00:07,  3.34it/s]
 68% 51/75 [00:13<00:06,  3.51it/s]
 69% 52/75 [00:13<00:06,  3.61it/s]
 71% 53/75 [00:13<00:06,  3.18it/s]
 72% 54/75 [00:14<00:06,  3.23it/s]
 73% 55/75 [00:14<00:05,  3.52it/s]
 75% 56/75 [00:14<00:05,  3.47it/s]
 76% 57/75 [00:15<00:05,  3.42it/s]
 77% 58/75 [00:15<00:04,  3.58it/s]
 79% 59/75 [00:15<00:04,  3.75it/s]
 80% 60/75 [00:15<00:04,  3.58it/s]
 81% 61/75 [00:16<00:04,  3.49it/s]
 83% 62/75 [00:16<00:03,  3.37it/s]
 84% 63/75 [00:16<00:03,  3.41it/s]
 85% 64/75 [00:17<00:03,  3.46it/s]
 87% 65/75 [00:17<00:02,  3.59it/s]
 88% 66/75 [00:17<00:02,  3.94it/s]
 89% 67/75 [00:17<00:01,  4.02it/s]
 91% 68/75 [00:17<00:01,  3.85it/s]
 92% 69/75 [00:18<00:01,  3.85it/s]
 93% 70/75 [00:18<00:01,  4.05it/s]
 95% 71/75 [00:18<00:01,  3.92it/s]
 96% 72/75 [00:18<00:00,  3.89it/s]
 97% 73/75 [00:19<00:00,  3.91it/s]
 99% 74/75 [00:19<00:00,  3.68it/s]
                                          
{'eval_loss': 1.2926925420761108, 'eval_runtime': 20.1146, 'eval_samples_per_second': 3.729, 'eval_steps_per_second': 3.729, 'epoch': 0.01}
  0% 105/22194 [09:41<19:43:58,  3.22s/it]
100% 75/75 [00:20<00:00,  3.63it/s]
{'loss': 1.0452, 'learning_rate': 0.00019999075885438494, 'epoch': 0.01}
{'loss': 1.1059, 'learning_rate': 0.00019999056533083262, 'epoch': 0.01}
{'loss': 1.1206, 'learning_rate': 0.00019999036980198172, 'epoch': 0.01}
{'loss': 1.6458, 'learning_rate': 0.0001999901722678361, 'epoch': 0.01}
{'loss': 0.7857, 'learning_rate': 0.00019998997272839975, 'epoch': 0.01}
  0% 110/22194 [09:46<17:04:57,  2.78s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07,  9.69it/s]
  4% 3/75 [00:00<00:10,  7.03it/s]
  5% 4/75 [00:00<00:13,  5.30it/s]
  7% 5/75 [00:01<00:16,  4.13it/s]
  8% 6/75 [00:01<00:16,  4.31it/s]
  9% 7/75 [00:01<00:17,  3.80it/s]
 11% 8/75 [00:01<00:18,  3.61it/s]
 12% 9/75 [00:02<00:17,  3.80it/s]
 13% 10/75 [00:02<00:16,  3.83it/s]
 15% 11/75 [00:02<00:15,  4.04it/s]
 16% 12/75 [00:02<00:16,  3.85it/s]
 17% 13/75 [00:03<00:16,  3.76it/s]
 19% 14/75 [00:03<00:15,  3.84it/s]
 20% 15/75 [00:03<00:15,  3.94it/s]
 21% 16/75 [00:03<00:16,  3.64it/s]
 23% 17/75 [00:04<00:15,  3.74it/s]
 24% 18/75 [00:04<00:14,  3.94it/s]
 25% 19/75 [00:04<00:14,  3.77it/s]
 27% 20/75 [00:05<00:15,  3.63it/s]
 28% 21/75 [00:05<00:14,  3.63it/s]
 29% 22/75 [00:05<00:14,  3.59it/s]
 31% 23/75 [00:05<00:15,  3.42it/s]
 32% 24/75 [00:06<00:14,  3.53it/s]
 33% 25/75 [00:06<00:12,  3.85it/s]
 35% 26/75 [00:06<00:12,  4.02it/s]
 36% 27/75 [00:06<00:12,  3.85it/s]
 37% 28/75 [00:07<00:12,  3.73it/s]
 39% 29/75 [00:07<00:12,  3.65it/s]
 40% 30/75 [00:07<00:13,  3.45it/s]
 41% 31/75 [00:08<00:12,  3.42it/s]
 43% 32/75 [00:08<00:11,  3.66it/s]
 44% 33/75 [00:08<00:11,  3.64it/s]auto

 45% 34/75 [00:08<00:11,  3.52it/s]
 47% 35/75 [00:09<00:11,  3.42it/s]
 48% 36/75 [00:09<00:11,  3.47it/s]
 49% 37/75 [00:09<00:10,  3.46it/s]
 51% 38/75 [00:10<00:10,  3.48it/s]
 52% 39/75 [00:10<00:10,  3.48it/s]
 53% 40/75 [00:10<00:09,  3.54it/s]
 55% 41/75 [00:10<00:09,  3.55it/s]
 56% 42/75 [00:11<00:09,  3.52it/s]
 57% 43/75 [00:11<00:08,  3.61it/s]
 59% 44/75 [00:11<00:08,  3.84it/s]
 60% 45/75 [00:11<00:08,  3.68it/s]
 61% 46/75 [00:12<00:08,  3.41it/s]auto

 63% 47/75 [00:12<00:08,  3.27it/s]
 64% 48/75 [00:12<00:08,  3.25it/s]
 65% 49/75 [00:13<00:07,  3.29it/s]
 67% 50/75 [00:13<00:07,  3.22it/s]
 68% 51/75 [00:13<00:07,  3.40it/s]
 69% 52/75 [00:14<00:06,  3.54it/s]
 71% 53/75 [00:14<00:07,  3.14it/s]
 72% 54/75 [00:14<00:06,  3.08it/s]
 73% 55/75 [00:15<00:05,  3.47it/s]
 75% 56/75 [00:15<00:05,  3.45it/s]
 76% 57/75 [00:15<00:05,  3.34it/s]
 77% 58/75 [00:15<00:04,  3.41it/s]
 79% 59/75 [00:16<00:04,  3.78it/s]
 80% 60/75 [00:16<00:04,  3.54it/s]
 81% 61/75 [00:16<00:04,  3.49it/s]
 83% 62/75 [00:17<00:03,  3.33it/s]
 84% 63/75 [00:17<00:03,  3.37it/s]
 85% 64/75 [00:17<00:03,  3.46it/s]
 87% 65/75 [00:17<00:02,  3.62it/s]
 88% 66/75 [00:18<00:02,  3.97it/s]
 89% 67/75 [00:18<00:01,  4.05it/s]
 91% 68/75 [00:18<00:01,  3.86it/s]
 92% 69/75 [00:18<00:01,  3.87it/s]
 93% 70/75 [00:19<00:01,  4.07it/s]
 95% 71/75 [00:19<00:01,  3.92it/s]
 96% 72/75 [00:19<00:00,  3.90it/s]
 97% 73/75 [00:19<00:00,  3.93it/s]
 99% 74/75 [00:20<00:00,  3.70it/s]
                                          
{'eval_loss': 1.2794123888015747, 'eval_runtime': 20.7359, 'eval_samples_per_second': 3.617, 'eval_steps_per_second': 3.617, 'epoch': 0.01}
  0% 110/22194 [10:07<17:04:57,  2.78s/it]
100% 75/75 [00:20<00:00,  3.65it/s]
{'loss': 0.6842, 'learning_rate': 0.00019998977118367662, 'epoch': 0.02}
{'loss': 1.16, 'learning_rate': 0.00019998956763367086, 'epoch': 0.02}
{'loss': 1.1332, 'learning_rate': 0.00019998936207838645, 'epoch': 0.02}
{'loss': 1.053, 'learning_rate': 0.00019998915451782754, 'epoch': 0.02}
{'loss': 1.4185, 'learning_rate': 0.0001999889449519983, 'epoch': 0.02}
  1% 115/22194 [10:18<19:31:53,  3.18s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.43it/s]
  5% 4/75 [00:00<00:12,  5.65it/s]
  7% 5/75 [00:00<00:15,  4.52it/s]
  8% 6/75 [00:01<00:14,  4.70it/s]
  9% 7/75 [00:01<00:16,  4.06it/s]
 11% 8/75 [00:01<00:17,  3.90it/s]
 12% 9/75 [00:02<00:16,  4.06it/s]
 13% 10/75 [00:02<00:15,  4.11it/s]
 15% 11/75 [00:02<00:15,  4.26it/s]
 16% 12/75 [00:02<00:15,  4.08it/s]
 17% 13/75 [00:03<00:15,  3.95it/s]
 19% 14/75 [00:03<00:15,  4.01it/s]
 20% 15/75 [00:03<00:14,  4.09it/s]
 21% 16/75 [00:03<00:15,  3.81it/s]
 23% 17/75 [00:04<00:14,  3.89it/s]
 24% 18/75 [00:04<00:13,  4.07it/s]
 25% 19/75 [00:04<00:14,  3.93it/s]
 27% 20/75 [00:04<00:14,  3.77it/s]
 28% 21/75 [00:05<00:14,  3.75it/s]
 29% 22/75 [00:05<00:14,  3.70it/s]
 31% 23/75 [00:05<00:14,  3.54it/s]
 32% 24/75 [00:05<00:13,  3.66it/s]
 33% 25/75 [00:06<00:12,  3.93it/s]
 35% 26/75 [00:06<00:11,  4.17it/s]
 36% 27/75 [00:06<00:12,  4.00it/s]
 37% 28/75 [00:06<00:12,  3.84it/s]
 39% 29/75 [00:07<00:12,  3.77it/s]
 40% 30/75 [00:07<00:12,  3.55it/s]
 41% 31/75 [00:07<00:12,  3.53it/s]
 43% 32/75 [00:07<00:11,  3.79it/s]
 44% 33/75 [00:08<00:11,  3.77it/s]
 45% 34/75 [00:08<00:11,  3.63it/s]
 47% 35/75 [00:08<00:11,  3.51it/s]
 48% 36/75 [00:09<00:11,  3.52it/s]
 49% 37/75 [00:09<00:10,  3.53it/s]
 51% 38/75 [00:09<00:10,  3.56it/s]
 52% 39/75 [00:09<00:10,  3.54it/s]
 53% 40/75 [00:10<00:09,  3.61it/s]
 55% 41/75 [00:10<00:09,  3.62it/s]
 56% 42/75 [00:10<00:09,  3.58it/s]
 57% 43/75 [00:11<00:08,  3.68it/s]
 59% 44/75 [00:11<00:07,  3.91it/s]
 60% 45/75 [00:11<00:08,  3.74it/s]
 61% 46/75 [00:11<00:08,  3.47it/s]
 63% 47/75 [00:12<00:08,  3.30it/s]
 64% 48/75 [00:12<00:08,  3.31it/s]
 65% 49/75 [00:12<00:07,  3.32it/s]
 67% 50/75 [00:13<00:07,  3.25it/s]
 68% 51/75 [00:13<00:06,  3.43it/s]
 69% 52/75 [00:13<00:06,  3.56it/s]
 71% 53/75 [00:14<00:07,  3.14it/s]
 72% 54/75 [00:14<00:06,  3.10it/s]
 73% 55/75 [00:14<00:05,  3.49it/s]
 75% 56/75 [00:14<00:05,  3.47it/s]
 76% 57/75 [00:15<00:05,  3.37it/s]
 77% 58/75 [00:15<00:04,  3.44it/s]
 79% 59/75 [00:15<00:04,  3.83it/s]
 80% 60/75 [00:16<00:04,  3.54it/s]
 81% 61/75 [00:16<00:03,  3.53it/s]
 83% 62/75 [00:16<00:03,  3.34it/s]
 84% 63/75 [00:16<00:03,  3.43it/s]
 85% 64/75 [00:17<00:03,  3.40it/s]
 87% 65/75 [00:17<00:02,  3.53it/s]
 88% 66/75 [00:17<00:02,  4.01it/s]
 89% 67/75 [00:17<00:01,  4.08it/s]
 91% 68/75 [00:18<00:01,  3.86it/s]
 92% 69/75 [00:18<00:01,  3.89it/s]
 93% 70/75 [00:18<00:01,  4.08it/s]
 95% 71/75 [00:18<00:01,  3.94it/s]
 96% 72/75 [00:19<00:00,  3.87it/s]
 97% 73/75 [00:19<00:00,  3.93it/s]
 99% 74/75 [00:19<00:00,  3.71it/s]
                                          
{'eval_loss': 1.2721059322357178, 'eval_runtime': 20.3404, 'eval_samples_per_second': 3.687, 'eval_steps_per_second': 3.687, 'epoch': 0.02}
  1% 115/22194 [10:39<19:31:53,  3.18s/it]
100% 75/75 [00:20<00:00,  3.64it/s]
{'loss': 0.7514, 'learning_rate': 0.00019998873338090295, 'epoch': 0.02}
{'loss': 1.4436, 'learning_rate': 0.00019998851980454574, 'epoch': 0.02}
{'loss': 1.5387, 'learning_rate': 0.0001999883042229309, 'epoch': 0.02}
{'loss': 0.997, 'learning_rate': 0.00019998808663606278, 'epoch': 0.02}
{'loss': 1.3263, 'learning_rate': 0.00019998786704394578, 'epoch': 0.02}
  1% 120/22194 [10:43<16:12:13,  2.64s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 10.90it/s]
  5% 4/75 [00:00<00:13,  5.25it/s]
  7% 5/75 [00:01<00:15,  4.44it/s]
  8% 6/75 [00:01<00:15,  4.43it/s]
  9% 7/75 [00:01<00:17,  3.84it/s]
 11% 8/75 [00:01<00:18,  3.68it/s]
 12% 9/75 [00:02<00:17,  3.87it/s]
 13% 10/75 [00:02<00:16,  3.91it/s]
 15% 11/75 [00:02<00:15,  4.06it/s]
 16% 12/75 [00:02<00:16,  3.90it/s]
 17% 13/75 [00:03<00:16,  3.82it/s]
 19% 14/75 [00:03<00:16,  3.75it/s]
 20% 15/75 [00:03<00:15,  3.97it/s]
 21% 16/75 [00:03<00:15,  3.69it/s]
 23% 17/75 [00:04<00:15,  3.66it/s]
 24% 18/75 [00:04<00:14,  4.01it/s]
 25% 19/75 [00:04<00:14,  3.80it/s]
 27% 20/75 [00:05<00:14,  3.69it/s]
 28% 21/75 [00:05<00:14,  3.68it/s]
 29% 22/75 [00:05<00:14,  3.60it/s]
 31% 23/75 [00:05<00:15,  3.44it/s]
 32% 24/75 [00:06<00:14,  3.49it/s]
 33% 25/75 [00:06<00:13,  3.78it/s]
 35% 26/75 [00:06<00:12,  4.07it/s]
 36% 27/75 [00:06<00:12,  3.91it/s]
 37% 28/75 [00:07<00:13,  3.53it/s]
 39% 29/75 [00:07<00:12,  3.79it/s]
 40% 30/75 [00:07<00:12,  3.55it/s]
 41% 31/75 [00:08<00:13,  3.36it/s]
 43% 32/75 [00:08<00:11,  3.76it/s]
 44% 33/75 [00:08<00:11,  3.75it/s]
 45% 34/75 [00:08<00:11,  3.59it/s]
 47% 35/75 [00:09<00:11,  3.40it/s]
 48% 36/75 [00:09<00:11,  3.50it/s]
 49% 37/75 [00:09<00:10,  3.53it/s]
 51% 38/75 [00:10<00:10,  3.53it/s]
 52% 39/75 [00:10<00:10,  3.53it/s]
 53% 40/75 [00:10<00:09,  3.57it/s]
 55% 41/75 [00:10<00:09,  3.57it/s]
 56% 42/75 [00:11<00:09,  3.58it/s]
 57% 43/75 [00:11<00:08,  3.66it/s]
 59% 44/75 [00:11<00:08,  3.83it/s]
 60% 45/75 [00:11<00:08,  3.68it/s]
 61% 46/75 [00:12<00:08,  3.43it/s]
 63% 47/75 [00:12<00:08,  3.26it/s]
 64% 48/75 [00:12<00:08,  3.25it/s]
 65% 49/75 [00:13<00:07,  3.29it/s]
 67% 50/75 [00:13<00:07,  3.26it/s]
 68% 51/75 [00:13<00:07,  3.41it/s]
 69% 52/75 [00:14<00:06,  3.51it/s]
 71% 53/75 [00:14<00:07,  3.12it/s]
 72% 54/75 [00:14<00:06,  3.18it/s]
 73% 55/75 [00:14<00:05,  3.46it/s]
 75% 56/75 [00:15<00:05,  3.42it/s]
 76% 57/75 [00:15<00:05,  3.40it/s]
 77% 58/75 [00:15<00:04,  3.57it/s]
 79% 59/75 [00:16<00:04,  3.74it/s]
 80% 60/75 [00:16<00:04,  3.57it/s]
 81% 61/75 [00:16<00:04,  3.48it/s]
 83% 62/75 [00:16<00:03,  3.36it/s]
 84% 63/75 [00:17<00:03,  3.41it/s]
 85% 64/75 [00:17<00:03,  3.47it/s]
 87% 65/75 [00:17<00:02,  3.59it/s]
 88% 66/75 [00:18<00:02,  3.90it/s]
 89% 67/75 [00:18<00:01,  4.02it/s]
 91% 68/75 [00:18<00:01,  3.86it/s]
 92% 69/75 [00:18<00:01,  3.88it/s]
 93% 70/75 [00:18<00:01,  4.05it/s]
 95% 71/75 [00:19<00:01,  3.95it/s]
 96% 72/75 [00:19<00:00,  3.74it/s]
 97% 73/75 [00:19<00:00,  4.02it/s]
 99% 74/75 [00:20<00:00,  3.76it/s]
                                          
{'eval_loss': 1.2644238471984863, 'eval_runtime': 20.6068, 'eval_samples_per_second': 3.64, 'eval_steps_per_second': 3.64, 'epoch': 0.02}
  1% 120/22194 [11:03<16:12:13,  2.64s/it]
100% 75/75 [00:20<00:00,  3.72it/s]
{'loss': 1.8826, 'learning_rate': 0.00019998764544658425, 'epoch': 0.02}
{'loss': 1.3692, 'learning_rate': 0.00019998742184398265, 'epoch': 0.02}
{'loss': 1.2665, 'learning_rate': 0.0001999871962361455, 'epoch': 0.02}
{'loss': 0.7873, 'learning_rate': 0.00019998696862307726, 'epoch': 0.02}
{'loss': 1.3467, 'learning_rate': 0.00019998673900478256, 'epoch': 0.02}
  1% 125/22194 [11:15<19:25:27,  3.17s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:07, 10.02it/s]
  5% 4/75 [00:00<00:13,  5.34it/s]
  7% 5/75 [00:00<00:15,  4.55it/s]
  8% 6/75 [00:01<00:15,  4.56it/s]
  9% 7/75 [00:01<00:17,  3.95it/s]
 11% 8/75 [00:01<00:17,  3.83it/s]
 12% 9/75 [00:02<00:16,  4.00it/s]
 13% 10/75 [00:02<00:16,  4.02it/s]
 15% 11/75 [00:02<00:15,  4.17it/s]
 16% 12/75 [00:02<00:15,  4.05it/s]
 17% 13/75 [00:03<00:15,  3.97it/s]
 19% 14/75 [00:03<00:15,  3.86it/s]
 20% 15/75 [00:03<00:14,  4.17it/s]
 21% 16/75 [00:03<00:15,  3.81it/s]
 23% 17/75 [00:04<00:15,  3.85it/s]
 24% 18/75 [00:04<00:13,  4.14it/s]
 25% 19/75 [00:04<00:14,  3.91it/s]
 27% 20/75 [00:04<00:14,  3.79it/s]
 28% 21/75 [00:05<00:14,  3.75it/s]
 29% 22/75 [00:05<00:14,  3.66it/s]
 31% 23/75 [00:05<00:14,  3.47it/s]
 32% 24/75 [00:06<00:14,  3.54it/s]
 33% 25/75 [00:06<00:12,  3.90it/s]
 35% 26/75 [00:06<00:11,  4.15it/s]
 36% 27/75 [00:06<00:12,  3.97it/s]
 37% 28/75 [00:06<00:12,  3.84it/s]
 39% 29/75 [00:07<00:12,  3.80it/s]
 40% 30/75 [00:07<00:12,  3.57it/s]
 41% 31/75 [00:07<00:12,  3.43it/s]
 43% 32/75 [00:08<00:11,  3.77it/s]
 44% 33/75 [00:08<00:11,  3.78it/s]
 45% 34/75 [00:08<00:11,  3.67it/s]
 47% 35/75 [00:08<00:11,  3.55it/s]
 48% 36/75 [00:09<00:11,  3.54it/s]
 49% 37/75 [00:09<00:10,  3.58it/s]
 51% 38/75 [00:09<00:10,  3.58it/s]
 52% 39/75 [00:10<00:10,  3.58it/s]
 53% 40/75 [00:10<00:09,  3.60it/s]
 55% 41/75 [00:10<00:09,  3.59it/s]
 56% 42/75 [00:10<00:09,  3.56it/s]
 57% 43/75 [00:11<00:08,  3.65it/s]
 59% 44/75 [00:11<00:07,  3.88it/s]
 60% 45/75 [00:11<00:08,  3.71it/s]
 61% 46/75 [00:12<00:08,  3.44it/s]
 63% 47/75 [00:12<00:08,  3.28it/s]
 64% 48/75 [00:12<00:08,  3.27it/s]
 65% 49/75 [00:12<00:07,  3.30it/s]
 67% 50/75 [00:13<00:07,  3.26it/s]
 68% 51/75 [00:13<00:07,  3.42it/s]
 69% 52/75 [00:13<00:06,  3.52it/s]
 71% 53/75 [00:14<00:07,  3.13it/s]
 72% 54/75 [00:14<00:06,  3.17it/s]
 73% 55/75 [00:14<00:05,  3.44it/s]
 75% 56/75 [00:15<00:05,  3.39it/s]
 76% 57/75 [00:15<00:05,  3.36it/s]
 77% 58/75 [00:15<00:04,  3.52it/s]
 79% 59/75 [00:15<00:04,  3.70it/s]
 80% 60/75 [00:16<00:04,  3.52it/s]
 81% 61/75 [00:16<00:04,  3.43it/s]
 83% 62/75 [00:16<00:03,  3.33it/s]
 84% 63/75 [00:17<00:03,  3.37it/s]
 85% 64/75 [00:17<00:03,  3.43it/s]
 87% 65/75 [00:17<00:02,  3.55it/s]
 88% 66/75 [00:17<00:02,  3.89it/s]
 89% 67/75 [00:18<00:02,  3.98it/s]
 91% 68/75 [00:18<00:01,  3.80it/s]
 92% 69/75 [00:18<00:01,  3.72it/s]
 93% 70/75 [00:18<00:01,  4.02it/s]
 95% 71/75 [00:19<00:01,  3.89it/s]
 96% 72/75 [00:19<00:00,  3.73it/s]
 97% 73/75 [00:19<00:00,  3.91it/s]
 99% 74/75 [00:19<00:00,  3.65it/s]
                                          
{'eval_loss': 1.273656964302063, 'eval_runtime': 20.4669, 'eval_samples_per_second': 3.664, 'eval_steps_per_second': 3.664, 'epoch': 0.02}
  1% 125/22194 [11:36<19:25:27,  3.17s/it]
100% 75/75 [00:20<00:00,  3.60it/s]
{'loss': 1.2297, 'learning_rate': 0.00019998650738126594, 'epoch': 0.02}
{'loss': 1.2587, 'learning_rate': 0.00019998627375253212, 'epoch': 0.02}
{'loss': 0.9847, 'learning_rate': 0.00019998603811858571, 'epoch': 0.02}
{'loss': 1.0413, 'learning_rate': 0.00019998580047943148, 'epoch': 0.02}
{'loss': 1.5374, 'learning_rate': 0.0001999855608350742, 'epoch': 0.02}
  1% 130/22194 [11:39<15:57:50,  2.60s/it]
  0% 0/75 [00:00<?, ?it/s]
  3% 2/75 [00:00<00:06, 11.81it/s]
  5% 4/75 [00:00<00:13,  5.37it/s]
  7% 5/75 [00:00<00:15,  4.52it/s]
  8% 6/75 [00:01<00:15,  4.38it/s]
  9% 7/75 [00:01<00:17,  3.84it/s]
 11% 8/75 [00:01<00:18,  3.63it/s]
 12% 9/75 [00:02<00:17,  3.83it/s]
 13% 10/75 [00:02<00:16,  3.86it/s]
 15% 11/75 [00:02<00:15,  4.01it/s]
 16% 12/75 [00:02<00:16,  3.84it/s]
 17% 13/75 [00:03<00:16,  3.77it/s]
 19% 14/75 [00:03<00:15,  3.82it/s]
 20% 15/75 [00:03<00:15,  3.87it/s]
 21% 16/75 [00:03<00:16,  3.62it/s]
 23% 17/75 [00:04<00:15,  3.73it/s]
 24% 18/75 [00:04<00:14,  3.91it/s]
 25% 19/75 [00:04<00:15,  3.71it/s]
 27% 20/75 [00:05<00:15,  3.61it/s]
 28% 21/75 [00:05<00:15,  3.60it/s]
 29% 22/75 [00:05<00:14,  3.55it/s]
 31% 23/75 [00:05<00:15,  3.41it/s]
 32% 24/75 [00:06<00:14,  3.53it/s]
 33% 25/75 [00:06<00:13,  3.83it/s]
 35% 26/75 [00:06<00:12,  3.96it/s]
 36% 27/75 [00:06<00:12,  3.80it/s]
 37% 28/75 [00:07<00:12,  3.71it/s]
 39% 29/75 [00:07<00:12,  3.64it/s]
 40% 30/75 [00:07<00:13,  3.42it/s]
 41% 31/75 [00:08<00:12,  3.43it/s]
 43% 32/75 [00:08<00:11,  3.64it/s]
 44% 33/75 [00:08<00:11,  3.64it/s]
 45% 34/75 [00:08<00:11,  3.50it/s]
 47% 35/75 [00:09<00:11,  3.42it/s]
 48% 36/75 [00:09<00:11,  3.45it/s]
 49% 37/75 [00:09<00:10,  3.48it/s]
 51% 38/75 [00:10<00:10,  3.49it/s]
 52% 39/75 [00:10<00:10,  3.46it/s]
 53% 40/75 [00:10<00:09,  3.55it/s]
 55% 41/75 [00:10<00:09,  3.53it/s]
 56% 42/75 [00:11<00:09,  3.52it/s]
 57% 43/75 [00:11<00:08,  3.57it/s]
 59% 44/75 [00:11<00:08,  3.86it/s]
 60% 45/75 [00:12<00:08,  3.69it/s]
 61% 46/75 [00:12<00:08,  3.44it/s]
 63% 47/75 [00:12<00:08,  3.26it/s]
 64% 48/75 [00:13<00:08,  3.17it/s]
 65% 49/75 [00:13<00:07,  3.31it/s]
 67% 50/75 [00:13<00:07,  3.23it/s]
 68% 51/75 [00:13<00:07,  3.32it/s]
 69% 52/75 [00:14<00:06,  3.57it/s]
 71% 53/75 [00:14<00:06,  3.16it/s]
 72% 54/75 [00:14<00:06,  3.09it/s]
 73% 55/75 [00:15<00:05,  3.44it/s]
 75% 56/75 [00:15<00:05,  3.51it/s]
 76% 57/75 [00:15<00:05,  3.35it/s]
 77% 58/75 [00:15<00:04,  3.49it/s]
 79% 59/75 [00:16<00:04,  3.84it/s]
 80% 60/75 [00:16<00:04,  3.66it/s]
 81% 61/75 [00:16<00:03,  3.52it/s]
 83% 62/75 [00:17<00:03,  3.39it/s]
 84% 63/75 [00:17<00:03,  3.43it/s]
 85% 64/75 [00:17<00:03,  3.48it/s]
 87% 65/75 [00:17<00:02,  3.60it/s]
 88% 66/75 [00:18<00:02,  3.91it/s]
 89% 67/75 [00:18<00:01,  4.02it/s]
 91% 68/75 [00:18<00:01,  3.87it/s]
 92% 69/75 [00:18<00:01,  3.88it/s]
 93% 70/75 [00:19<00:01,  4.05it/s]
 95% 71/75 [00:19<00:01,  3.93it/s]
 96% 72/75 [00:19<00:00,  3.91it/s]
 97% 73/75 [00:19<00:00,  3.97it/s]
 99% 74/75 [00:20<00:00,  3.74it/s]
                                          
{'eval_loss': 1.2697356939315796, 'eval_runtime': 20.7509, 'eval_samples_per_second': 3.614, 'eval_steps_per_second': 3.614, 'epoch': 0.02}
  1% 130/22194 [12:00<15:57:50,  2.60s/it]
100% 75/75 [00:20<00:00,  3.68it/s]
                                   

%env ACCELERATE_USE_DEEPSPEED=false

env: ACCELERATE_USE_DEEPSPEED=false

!python scripts/ft.py examples/falcon/config-1b-qlora.yml --inference

===================================BUG REPORT===================================
Welcome to bitsandbytes. For bug reports, please run

python -m bitsandbytes

 and submit this information together with your error trace to: https://github.com/TimDettmers/bitsandbytes/issues
================================================================================
bin /usr/local/lib/python3.10/dist-packages/bitsandbytes/libbitsandbytes_cuda118.so
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: /usr/lib64-nvidia did not contain ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] as expected! Searching further paths...
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('/sys/fs/cgroup/memory.events /var/colab/cgroup/jupyter-children/memory.events')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('8013'), PosixPath('http'), PosixPath('//172.28.0.1')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('--logtostderr --listen_host=172.28.0.12 --target_host=172.28.0.12 --tunnel_background_save_url=https'), PosixPath('//colab.research.google.com/tun/m/cc48301118ce562b961b3c22d803539adc1e0c19/gpu-t4-s-tccs3c5dv19r --tunnel_background_save_delay=10s --tunnel_periodic_background_save_frequency=30m0s --enable_output_coalescing=true --output_coalescing_required=true')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('/env/python')}
  warn(msg)
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('//ipykernel.pylab.backend_inline'), PosixPath('module')}
  warn(msg)
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching in backup paths...
/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: Found duplicate ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] files: {PosixPath('/usr/local/cuda/lib64/libcudart.so'), PosixPath('/usr/local/cuda/lib64/libcudart.so.11.0')}.. We'll flip a coin and try one of these, in order to fail forward.
Either way, this might cause trouble in the future:
If you get `CUDA error: invalid device function` errors, the above might be the cause and the solution is to make sure only one ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] in the paths that we search based on your env.
  warn(msg)
CUDA SETUP: CUDA runtime path found: /usr/local/cuda/lib64/libcudart.so
CUDA SETUP: Highest compute capability among GPUs detected: 7.5
CUDA SETUP: Detected CUDA version 118
CUDA SETUP: Loading binary /usr/local/lib/python3.10/dist-packages/bitsandbytes/libbitsandbytes_cuda118.so...
[2023-06-10 13:51:29,095] [INFO] [real_accelerator.py:110:get_accelerator] Setting ds_accelerator to cuda (auto detect)
2023-06-10 13:51:31.276433: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
INFO:root:train_ex before
WARNING:root:`trust_remote_code` is set to true. Please make sure that you reviewed the remote code/model.
INFO:root:loading tokenizer... tiiuae/falcon-rw-1b
Using pad_token, but it is not set yet.
INFO:root:loading model and peft_config...
INFO:root:converting PEFT model w/ prepare_model_for_kbit_training
INFO:root:found linear modules: ['query_key_value', 'dense_h_to_4h', 'dense', 'dense_4h_to_h']
trainable params: 50,331,648 || all params: 757,911,552 || trainable%: 6.6408339953630895
INFO:root:calling do_inference function
Give me an instruction (Ctrl + D to finish): 
Write a passage with no less than 500 words that starts with "I love rainy days"
Error in sys.excepthook:
Traceback (most recent call last):
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1699, in print
    extend(render(renderable, render_options))
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1335, in render
    yield from self.render(render_output, _options)
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1331, in render
    for render_output in iter_render:
  File "/usr/local/lib/python3.10/dist-packages/rich/constrain.py", line 29, in __rich_console__
    yield from console.render(self.renderable, child_options)
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1331, in render
    for render_output in iter_render:
  File "/usr/local/lib/python3.10/dist-packages/rich/panel.py", line 220, in __rich_console__
    lines = console.render_lines(renderable, child_options, style=style)
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1371, in render_lines
    lines = list(
  File "/usr/local/lib/python3.10/dist-packages/rich/segment.py", line 292, in split_and_crop_lines
    for segment in segments:
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1331, in render
    for render_output in iter_render:
  File "/usr/local/lib/python3.10/dist-packages/rich/padding.py", line 97, in __rich_console__
    lines = console.render_lines(
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1371, in render_lines
    lines = list(
  File "/usr/local/lib/python3.10/dist-packages/rich/segment.py", line 292, in split_and_crop_lines
    for segment in segments:
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1335, in render
    yield from self.render(render_output, _options)
  File "/usr/local/lib/python3.10/dist-packages/rich/console.py", line 1331, in render
    for render_output in iter_render:
  File "/usr/local/lib/python3.10/dist-packages/rich/syntax.py", line 612, in __rich_console__
    segments = Segments(self._get_syntax(console, options))
  File "/usr/local/lib/python3.10/dist-packages/rich/segment.py", line 668, in __init__
    self.segments = list(segments)
  File "/usr/local/lib/python3.10/dist-packages/rich/syntax.py", line 640, in _get_syntax
    text = self.highlight(processed_code, self.line_range)
  File "/usr/local/lib/python3.10/dist-packages/rich/syntax.py", line 470, in highlight
    lexer = self.lexer
  File "/usr/local/lib/python3.10/dist-packages/rich/syntax.py", line 433, in lexer
    return get_lexer_by_name(
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexers/__init__.py", line 106, in get_lexer_by_name
    return _lexer_cache[name](**options)
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexer.py", line 581, in __call__
    cls._tokens = cls.process_tokendef('', cls.get_tokendefs())
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexer.py", line 520, in process_tokendef
    cls._process_state(tokendefs, processed, state)
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexer.py", line 483, in _process_state
    tokens.extend(cls._process_state(unprocessed, processed,
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexer.py", line 499, in _process_state
    rex = cls._process_regex(tdef[0], rflags, state)
  File "/usr/local/lib/python3.10/dist-packages/pygments/lexer.py", line 428, in _process_regex
    return re.compile(regex, rflags).match
  File "/usr/lib/python3.10/re.py", line 251, in compile
    return _compile(pattern, flags)
  File "/usr/lib/python3.10/re.py", line 303, in _compile
    p = sre_compile.compile(pattern, flags)
  File "/usr/lib/python3.10/sre_compile.py", line 792, in compile
    code = _code(p, flags)
  File "/usr/lib/python3.10/sre_compile.py", line 628, in _code
    _compile_info(code, p, flags)
  File "/usr/lib/python3.10/sre_compile.py", line 564, in _compile_info
    lo, hi = pattern.getwidth()
  File "/usr/lib/python3.10/sre_parse.py", line 181, in getwidth
    if op is BRANCH:
KeyboardInterrupt

Original exception was:
Traceback (most recent call last):
  File "/content/axolotl/scripts/ft.py", line 29, in <module>
    fire.Fire(train_ex)
  File "/usr/local/lib/python3.10/dist-packages/fire/core.py", line 141, in Fire
    component_trace = _Fire(component, args, parsed_flag_args, context, name)
  File "/usr/local/lib/python3.10/dist-packages/fire/core.py", line 475, in _Fire
    component, remaining_args = _CallAndUpdateTrace(
  File "/usr/local/lib/python3.10/dist-packages/fire/core.py", line 691, in _CallAndUpdateTrace
    component = fn(*varargs, **kwargs)
  File "/content/axolotl/scripts/ft.py", line 16, in train_ex
    finetune.train(config, prepare_ds_only, **kwargs)
  File "/content/axolotl/scripts/finetune.py", line 238, in train
    do_inference(cfg, model, tokenizer)
  File "/content/axolotl/scripts/finetune.py", line 75, in do_inference
    instruction = get_multi_line_input()
  File "/content/axolotl/scripts/finetune.py", line 60, in get_multi_line_input
    for line in sys.stdin:
KeyboardInterrupt
^C

!pip install colab-xterm
%load_ext colabxterm

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting colab-xterm
  Downloading colab_xterm-0.1.2-py3-none-any.whl (115 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 0.0/115.3 kB ? eta -:--:--
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 115.3/115.3 kB 4.3 MB/s eta 0:00:00
Requirement already satisfied: ptyprocess~=0.7.0 in /usr/local/lib/python3.10/dist-packages (from colab-xterm) (0.7.0)
Requirement already satisfied: tornado>5.1 in /usr/local/lib/python3.10/dist-packages (from colab-xterm) (6.3.1)
Installing collected packages: colab-xterm
Successfully installed colab-xterm-0.1.2

%%writefile input.txt
Hi

Writing input.txt

!pip install pexpect

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Requirement already satisfied: pexpect in /usr/local/lib/python3.10/dist-packages (4.8.0)
Requirement already satisfied: ptyprocess>=0.5 in /usr/local/lib/python3.10/dist-packages (from pexpect) (0.7.0)

import pexpect
child = pexpect.spawn('python scripts/ft.py examples/falcon/config-1b-qlora.yml --inference')

child.isalive()

False

child.sendintr()

child.read(5)

---------------------------------------------------------------------------
TIMEOUT                                   Traceback (most recent call last)
<ipython-input-171-bbdb3db4cc80> in <cell line: 1>()
----> 1 child.read(5)

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in read(self, size)
    454         cre = re.compile(self._coerce_expect_string('.{%d}' % size), re.DOTALL)
    455         # delimiter default is EOF
--> 456         index = self.expect([cre, self.delimiter])
    457         if index == 0:
    458             ### FIXME self.before should be ''. Should I assert this?

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in expect(self, pattern, timeout, searchwindowsize, async_, **kw)
    341 
    342         compiled_pattern_list = self.compile_pattern_list(pattern)
--> 343         return self.expect_list(compiled_pattern_list,
    344                 timeout, searchwindowsize, async_)
    345 

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in expect_list(self, pattern_list, timeout, searchwindowsize, async_, **kw)
    370             return expect_async(exp, timeout)
    371         else:
--> 372             return exp.expect_loop(timeout)
    373 
    374     def expect_exact(self, pattern_list, timeout=-1, searchwindowsize=-1,

/usr/local/lib/python3.10/dist-packages/pexpect/expect.py in expect_loop(self, timeout)
    179             return self.eof(e)
    180         except TIMEOUT as e:
--> 181             return self.timeout(e)
    182         except:
    183             self.errored()

/usr/local/lib/python3.10/dist-packages/pexpect/expect.py in timeout(self, err)
    142             exc = TIMEOUT(msg)
    143             exc.__cause__ = None    # in Python 3.x we can use "raise exc from None"
--> 144             raise exc
    145 
    146     def errored(self):

TIMEOUT: Timeout exceeded.
<pexpect.pty_spawn.spawn object at 0x7f7d5886c8e0>
command: /usr/local/bin/python
args: ['/usr/local/bin/python', 'scripts/ft.py', 'examples/falcon/config-1b-qlora.yml', '--inference']
buffer (last 100 chars): b'u?\r\n'
before (last 100 chars): b'u?\r\n'
after: <class 'pexpect.exceptions.TIMEOUT'>
match: None
match_index: None
exitstatus: None
flag_eof: False
pid: 57403
child_fd: 49
closed: False
timeout: 30
delimiter: <class 'pexpect.exceptions.EOF'>
logfile: None
logfile_read: None
logfile_send: None
maxread: 2000
ignorecase: False
searchwindowsize: None
delaybeforesend: 0.05
delayafterclose: 0.1
delayafterterminate: 0.1
searcher: searcher_re:
    0: re.compile(b'.{5}')
    1: EOF

child.sendline("Who are you?")

13

child.sendeof()

child.sendcontrol('d')

1

child.read(300)

b"ar/colab/cgroup/jupyter-children/memory.events')}\r\n  warn(msg)\r\n/usr/local/lib/python3.10/dist-packages/bitsandbytes/cuda_setup/main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {PosixPath('8013'), PosixPath('http'), PosixPath('//172.28.0"

child.interact()

while child.isalive():
  print(child.readline())

---------------------------------------------------------------------------
TIMEOUT                                   Traceback (most recent call last)
<ipython-input-202-780483499b14> in <cell line: 1>()
      1 while child.isalive():
----> 2   print(child.readline())

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in readline(self, size)
    475             return self.string_type()
    476         # delimiter default is EOF
--> 477         index = self.expect([self.crlf, self.delimiter])
    478         if index == 0:
    479             return self.before + self.crlf

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in expect(self, pattern, timeout, searchwindowsize, async_, **kw)
    341 
    342         compiled_pattern_list = self.compile_pattern_list(pattern)
--> 343         return self.expect_list(compiled_pattern_list,
    344                 timeout, searchwindowsize, async_)
    345 

/usr/local/lib/python3.10/dist-packages/pexpect/spawnbase.py in expect_list(self, pattern_list, timeout, searchwindowsize, async_, **kw)
    370             return expect_async(exp, timeout)
    371         else:
--> 372             return exp.expect_loop(timeout)
    373 
    374     def expect_exact(self, pattern_list, timeout=-1, searchwindowsize=-1,

/usr/local/lib/python3.10/dist-packages/pexpect/expect.py in expect_loop(self, timeout)
    179             return self.eof(e)
    180         except TIMEOUT as e:
--> 181             return self.timeout(e)
    182         except:
    183             self.errored()

/usr/local/lib/python3.10/dist-packages/pexpect/expect.py in timeout(self, err)
    142             exc = TIMEOUT(msg)
    143             exc.__cause__ = None    # in Python 3.x we can use "raise exc from None"
--> 144             raise exc
    145 
    146     def errored(self):

TIMEOUT: Timeout exceeded.
<pexpect.pty_spawn.spawn object at 0x7f7d587c5b40>
command: /usr/local/bin/python
args: ['/usr/local/bin/python', 'scripts/ft.py', 'examples/falcon/config-1b-qlora.yml', '--inference']
buffer (last 100 chars): b''
before (last 100 chars): ''
after: <class 'pexpect.exceptions.TIMEOUT'>
match: None
match_index: None
exitstatus: None
flag_eof: False
pid: 60024
child_fd: 50
closed: False
timeout: 30
delimiter: <class 'pexpect.exceptions.EOF'>
logfile: None
logfile_read: None
logfile_send: None
maxread: 2000
ignorecase: False
searchwindowsize: None
delaybeforesend: 0.05
delayafterclose: 0.1
delayafterterminate: 0.1
searcher: searcher_re:
    0: re.compile(b'\r\n')
    1: EOF

from subprocess import Popen, PIPE, STDOUT

p = Popen(['python', 'scripts/ft.py', 'examples/falcon/config-1b-qlora.yml', '--inference'], stdin=PIPE, stdout=PIPE, stderr=STDOUT)    
p.stdin.write('Hi'.encode())
p.stdin.flush()
p.stdout.readline()
p.wait()

---------------------------------------------------------------------------
KeyboardInterrupt                         Traceback (most recent call last)
<ipython-input-79-3993eb351ac5> in <cell line: 7>()
      5 p.stdin.flush()
      6 
----> 7 p.wait()

/usr/lib/python3.10/subprocess.py in wait(self, timeout)
   1207             endtime = _time() + timeout
   1208         try:
-> 1209             return self._wait(timeout=timeout)
   1210         except KeyboardInterrupt:
   1211             # https://bugs.python.org/issue25942

/usr/lib/python3.10/subprocess.py in _wait(self, timeout)
   1957                         if self.returncode is not None:
   1958                             break  # Another thread waited.
-> 1959                         (pid, sts) = self._try_wait(0)
   1960                         # Check the pid and loop as waitpid has been known to
   1961                         # return 0 even without WNOHANG in odd situations.

/usr/lib/python3.10/subprocess.py in _try_wait(self, wait_flags)
   1915             """All callers to this function MUST hold self._waitpid_lock."""
   1916             try:
-> 1917                 (pid, sts) = os.waitpid(self.pid, wait_flags)
   1918             except ChildProcessError:
   1919                 # This happens if SIGCLD is set to be ignored or waiting

KeyboardInterrupt: 

!pip install nvitop

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Collecting nvitop
  Downloading nvitop-1.1.2-py3-none-any.whl (206 kB)
     ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 206.8/206.8 kB 5.5 MB/s eta 0:00:00
Collecting nvidia-ml-py<11.526.0a0,>=11.450.51 (from nvitop)
  Downloading nvidia_ml_py-11.525.112-py3-none-any.whl (35 kB)
Requirement already satisfied: psutil>=5.6.6 in /usr/local/lib/python3.10/dist-packages (from nvitop) (5.9.5)
Requirement already satisfied: cachetools>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from nvitop) (5.3.0)
Requirement already satisfied: termcolor>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from nvitop) (2.3.0)
Installing collected packages: nvidia-ml-py, nvitop
Successfully installed nvidia-ml-py-11.525.112 nvitop-1.1.2

%xterm

Launching Xterm...

!rm -rf ./qlora-out

!accelerate config --config_file configs/accelerate/default_config.yaml default

[2023-06-10 13:42:53,481] [INFO] [real_accelerator.py:110:get_accelerator] Setting ds_accelerator to cuda (auto detect)
2023-06-10 13:42:57.209334: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT
Configuration already exists at /root/.cache/huggingface/accelerate/default_config.yaml, will not override. Run `accelerate config` manually or pass a different `save_location`.

!cat /root/.cache/huggingface/accelerate/default_config.yaml

{
  "compute_environment": "LOCAL_MACHINE",
  "distributed_type": "NO",
  "downcast_bf16": false,
  "machine_rank": 0,
  "main_training_function": "main",
  "mixed_precision": "no",
  "num_machines": 1,
  "num_processes": 1,
  "rdzv_backend": "static",
  "same_network": false,
  "tpu_use_cluster": false,
  "tpu_use_sudo": false,
  "use_cpu": false
}