Hugging Face's logo

tangledgroup
/
tangled-alpha-0.1-core

Text Generation
Transformers
chat
core
base
instruct
reason
conversational
Model card Files Community
tangled-alpha-0.1-core / README.md
mtasic85's picture
mtasic85
micro_batch_size: 4
50de401
preview code
|
raw
history blame
5.28 kB
metadata
license: mit
pipeline_tag: text-generation
library_name: transformers
language:
  - en
  - am
  - ar
  - as
  - az
  - be
  - bg
  - bn
  - br
  - bs
  - ca
  - cs
  - cy
  - da
  - de
  - el
  - eo
  - es
  - et
  - eu
  - fa
  - ff
  - fi
  - fr
  - fy
  - ga
  - gd
  - gl
  - gn
  - gu
  - ha
  - he
  - hi
  - hr
  - ht
  - hu
  - hy
  - id
  - ig
  - is
  - it
  - ja
  - jv
  - ka
  - kk
  - km
  - kn
  - ko
  - ku
  - ky
  - la
  - lg
  - li
  - ln
  - lo
  - lt
  - lv
  - mg
  - mk
  - ml
  - mn
  - mr
  - ms
  - my
  - ne
  - nl
  - 'no'
  - ns
  - om
  - or
  - pa
  - pl
  - ps
  - pt
  - qu
  - rm
  - ro
  - ru
  - sa
  - si
  - sc
  - sd
  - sk
  - sl
  - so
  - sq
  - sr
  - ss
  - su
  - sv
  - sw
  - ta
  - te
  - th
  - tl
  - tn
  - tr
  - ug
  - uk
  - ur
  - uz
  - vi
  - wo
  - xh
  - yi
  - yo
  - zu
datasets:
  - ontocord/fineweb-permissive-multilingual-2m
  - distily/c4_multilingual_1M
  - data-silence/sumnews
  - xu-song/cc100-samples
  - badrex/llm-emoji-dataset
  - fblgit/simple-math
  - Gusarich/math-expressions-1m
  - neuralwork/arxiver
  - christopher/rosetta-code
  - nampdn-ai/tiny-codes
  - JeanKaddour/minipile
  - NousResearch/hermes-function-calling-v1
  - simplescaling/s1K-1.1
  - mlabonne/open-perfectblend
  - allenai/tulu-3-sft-mixture
  - rombodawg/Everything_Instruct_Multilingual
  - open-r1/OpenR1-Math-220k
  - open-thoughts/OpenThoughts-114k
  - cognitivecomputations/dolphin-r1
  - simplescaling/s1K-1.1
tags:
  - chat
  - core
  - base
  - instruct
  - reason

tangled-alpha-0.1-core

logo

time python -B prepare_core_datasets.py

Progress: 100%|████████| 220/220 [23:15<00:00,  6.34s/it]
Workers are finished.██| 220/220 [23:15<00:00,  6.34s/it]
Finished data processing!
i=0, block_size=8192, chunk_size=16384000, len(dataset)=893355, len(dataset) * block_size=7318364160
Total number of tokens in the optimized dataset '../core-data-0-8192-2000' is 7318364160

CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt pretrain --config pretrain-core-model.yaml

Seed set to 23
Time to instantiate model: 0.23 seconds.
Total parameters: 138,084,864
Verifying settings ...
Measured TFLOPs: 6972.54
Epoch 1 | iter 256 step 1 | loss train: 10.530, val: n/a | iter time: 3230.47 ms (step) remaining time: 3 days, 5:34:13
Epoch 1 | iter 512 step 2 | loss train: 10.520, val: n/a | iter time: 589.19 ms (step) remaining time: 3 days, 0:40:40
Epoch 1 | iter 768 step 3 | loss train: 10.485, val: n/a | iter time: 591.81 ms (step) remaining time: 2 days, 23:01:54
Epoch 1 | iter 1024 step 4 | loss train: 10.447, val: n/a | iter time: 589.35 ms (step) remaining time: 2 days, 22:11:32
Epoch 1 | iter 1280 step 5 | loss train: 10.350, val: n/a | iter time: 589.38 ms (step) remaining time: 2 days, 21:40:13
Epoch 1 | iter 1536 step 6 | loss train: 10.241, val: n/a | iter time: 593.75 ms (step) remaining time: 2 days, 21:18:19
Epoch 1 | iter 1792 step 7 | loss train: 10.134, val: n/a | iter time: 592.92 ms (step) remaining time: 2 days, 21:01:58
Epoch 1 | iter 2048 step 8 | loss train: 10.049, val: n/a | iter time: 590.74 ms (step) remaining time: 2 days, 20:49:12
Epoch 1 | iter 2304 step 9 | loss train: 9.869, val: n/a | iter time: 594.27 ms (step) remaining time: 2 days, 20:39:10
Epoch 1 | iter 2560 step 10 | loss train: 9.771, val: n/a | iter time: 590.04 ms (step) remaining time: 2 days, 20:30:14
Epoch 1 | iter 2816 step 11 | loss train: 9.643, val: n/a | iter time: 588.32 ms (step) remaining time: 2 days, 20:22:22
Epoch 1 | iter 3072 step 12 | loss train: 9.557, val: n/a | iter time: 588.95 ms (step) remaining time: 2 days, 20:15:26
Epoch 1 | iter 3328 step 13 | loss train: 9.487, val: n/a | iter time: 589.32 ms (step) remaining time: 2 days, 20:09:05
Epoch 1 | iter 3584 step 14 | loss train: 9.413, val: n/a | iter time: 588.95 ms (step) remaining time: 2 days, 20:03:24
Epoch 1 | iter 3840 step 15 | loss train: 9.322, val: n/a | iter time: 591.62 ms (step) remaining time: 2 days, 19:58:18
Epoch 1 | iter 4096 step 16 | loss train: 9.241, val: n/a | iter time: 593.65 ms (step) remaining time: 2 days, 19:53:30
Epoch 1 | iter 4352 step 17 | loss train: 9.163, val: n/a | iter time: 593.89 ms (step) remaining time: 2 days, 19:49:00
Epoch 1 | iter 4608 step 18 | loss train: 9.122, val: n/a | iter time: 590.63 ms (step) remaining time: 2 days, 19:44:42
Epoch 1 | iter 4864 step 19 | loss train: 9.077, val: n/a | iter time: 590.87 ms (step) remaining time: 2 days, 19:40:47
Epoch 1 | iter 5120 step 20 | loss train: 9.018, val: n/a | iter time: 588.44 ms (step) remaining time: 2 days, 19:36:59
# ...

Backup wandb:

mv wandb wandb-pretrain-core

Chat with model:

CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True litgpt chat ../out/pretrain-core/final

CUDA_VISIBLE_DEVICES=0 CUDA_LAUNCH_BLOCKING=0 PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True time litgpt evaluate --tasks 'leaderboard' --out_dir '../evaluate/pretrain-core/leaderboard/' --batch_size 1 --dtype 'bfloat16' '../out/pretrain-core/final'

# ...