107 Commits

Author SHA1 Message Date
Sebastian Raschka
092b5b5429 topk comment 2024-08-20 20:44:15 -05:00
rasbt
f4e45a3f40 add note about duplicated cell 2024-08-19 21:04:18 -05:00
Sebastian Raschka
11e2f56af5 Note about MPS devices (#329) 2024-08-19 20:58:45 -05:00
Sebastian Raschka
0991c1ff24 Note about ch05 mps support (#324) 2024-08-19 07:40:24 -05:00
rasbt
742525cb28 remove redundant indentation 2024-08-16 07:54:02 -05:00
rasbt
6533ce63c1 fix code cell ordering 2024-08-12 19:04:05 -05:00
Sebastian Raschka
8d79fb13b0 Update README.md 2024-08-10 07:54:51 -05:00
Daniel Kleine
c91999b9f4 fixed bash command (#305)
Co-authored-by: Sebastian Raschka <mail@sebastianraschka.com>
2024-08-09 21:29:04 -05:00
TITC
3067ed83dc remove all non-English texts and notice (#304)
* remove all non-English texts and notice

1. almost 18GB txt left after `is_english` filtered.
2. remove notice use gutenberg's strip_headers
3. after re-run get_data.py, seems all data are under `gutenberg/data/.mirror` folder.

* some improvements

* update readme

---------

Co-authored-by: rasbt <mail@sebastianraschka.com>
2024-08-09 17:09:14 -05:00
TITC
7374d617b4 total training iters may equal to warmup_iters (#301)
total_training_iters=20, warmup_iters=20= len(train_loader) 4 multiply n_epochs 5, then ZeroDivisionError occurred.
```shell
Traceback (most recent call last):                                                                                                                                                                                                                                                                                              
  File "LLMs-from-scratch/ch05/05_bonus_hparam_tuning/hparam_search.py", line 191, in <module>                                             
    train_loss, val_loss = train_model(                                                                                                                                                                                                                                                                                         
                           ^^^^^^^^^^^^                                                                                                                         
  File "/mnt/raid1/docker/ai/LLMs-from-scratch/ch05/05_bonus_hparam_tuning/hparam_search.py", line 90, in train_model                                                                                                                                                                                                           
    progress = (global_step - warmup_iters) / (total_training_iters - warmup_iters)                                                                             
               ~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~                                                                                                                                                                                                                                             
ZeroDivisionError: division by zero 
```
2024-08-06 07:10:05 -05:00
SSebo
22681878a8 Update ch05.ipynb (#297)
typo
2024-08-05 07:12:27 -05:00
Sebastian Raschka
6dd8666d9c Test code in pytorch 2.4 (#285)
* test code in pytorch 2.4

* update
2024-07-24 21:53:41 -05:00
TITC
bce3a708f9 47,678-->48,725 (#281) 2024-07-22 21:24:57 -05:00
Sebastian Raschka
d0f3b034d8 Add download help message (#274) 2024-07-19 08:29:29 -05:00
rasbt
5e24a042c1 add links to summary sections 2024-06-29 07:33:26 -05:00
rasbt
0f43890a15 refresh cross entropy figure 2024-06-29 07:22:23 -05:00
Daniel Kleine
fb4e37ae15 fixed minor issues (#252)
* fixed typo

* fixed var name in md text
2024-06-29 06:38:25 -05:00
Daniel Kleine
7a54d383e7 minor fixes (#246)
* removed duplicated white spaces

* Update ch07/01_main-chapter-code/ch07.ipynb

* Update ch07/05_dataset-generation/llama3-ollama.ipynb

* removed duplicated white spaces

* fixed title again

---------

Co-authored-by: Sebastian Raschka <mail@sebastianraschka.com>
2024-06-25 17:30:30 -05:00
Sebastian Raschka
def84a039c Show epochs as integers on x-axis (#241)
* Show epochs as integers on x-axis

* Update ch07/01_main-chapter-code/previous_chapters.py

* remove extra s

* modify exercise plots

* update chapter 7 plot

* resave ch07 for better file diff
2024-06-23 07:41:25 -05:00
rasbt
0026e6206b update generate to match output in main chapter 2024-06-22 12:01:51 -05:00
Daniel Kleine
7e0c5c0975 minor fixes (#235)
* removed unnecessary imports

* removed unnecessary semicolons

* format markdown

* format markdown

* fixed markdown
2024-06-21 08:40:54 -05:00
rasbt
e1046746e8 remove redundant line 2024-06-20 10:12:28 -05:00
rasbt
cb194fa8fa fix device loading 2024-06-20 08:07:00 -05:00
rasbt
c1f9361428 add main and optional sections 2024-06-19 17:48:25 -05:00
rasbt
eb1da36e98 note about dropout 2024-06-19 17:37:48 -05:00
Daniel Kleine
73be1c592f fixed num_workers (#229)
* fixed num_workers

* ch06 & ch07: added num_workers to create_dataloader_v1
2024-06-19 17:36:46 -05:00
Sebastian Raschka
fcf8bcab0d Remove duplicated cell (#212)
* add a suggestion since code snippet has been repeated.

* remove duplicated cell

---------

Co-authored-by: Shuyib <benmainye@gmail.com>
2024-06-15 12:48:34 -05:00
rasbt
a796b9d657 explain truncation in ch05 2024-06-12 19:50:11 -05:00
Sebastian Raschka
8d3e58ff81 check gpt files (#208) 2024-06-12 07:19:10 -05:00
Daniel Kleine
e5c3c5ce99 minor bug fixes (#207)
* fixed path arg for create_dataset_csvs()

* updated assign_check() to remove user warning
2024-06-12 06:27:56 -05:00
rasbt
b2ff989174 distinguish better between main chapter code and bonus materials 2024-06-11 21:07:42 -05:00
Daniel Kleine
79210eb393 fixes for code (#206)
* updated .gitignore

* removed unused GELU import

* fixed model_configs, fixed all tensors on same device

* removed unused tiktoken

* update

* update hparam search

* remove redundant tokenizer argument

---------

Co-authored-by: rasbt <mail@sebastianraschka.com>
2024-06-11 20:59:48 -05:00
rasbt
f0e4c99bc3 fix typo in comment 2024-06-09 06:14:02 -05:00
Sebastian Raschka
40ba3a4068 Remove leftover instances of self.tokenizer (#201)
* Remove leftover instances of self.tokenizer

* add endoftext token
2024-06-08 14:57:34 -05:00
rasbt
5a1e0eecce fix learning rate scheduler 2024-06-03 07:06:42 -05:00
rasbt
f7e528fca6 update loss 2024-05-31 07:30:57 -05:00
Kumar Utsav
b48d436bfc Update ch05.ipynb
Fixed incorrect token ids
2024-05-29 20:34:23 +05:30
Sebastian Raschka
a0b5603423 Make header more clear 2024-05-25 10:44:12 -05:00
rasbt
fe8bb9291e update formatting 2024-05-24 07:20:37 -05:00
rasbt
aa084656e0 update how to retrieve learning rate 2024-05-23 17:19:01 -05:00
Daniel Kleine
69da9ed447 removed unnecessary .gitignore 2024-05-21 19:25:16 +00:00
rasbt
bc5cbbf1bd change defaults to 0 temp 2024-05-19 09:04:49 -05:00
rasbt
59f5ed8d68 use default value for temperature 2024-05-19 08:48:10 -05:00
rasbt
9d84935b69 add eos_id option for ch07 2024-05-18 12:35:40 -05:00
Daniel Kleine
e6012b944e fixed empty space 2024-05-17 10:44:18 +02:00
rasbt
b350daaa93 add readme 2024-05-13 08:50:55 -05:00
speed
7b34833ee1 fix 1024 characters to 1024 tokens (#152) 2024-05-11 13:17:07 -05:00
rasbt
bb59cbc525 link formatting 2024-04-30 06:26:23 -05:00
Sebastian Raschka
a5b353667d Rename drop_resid to drop_shortcut (#136) 2024-04-28 14:31:27 -05:00
rasbt
4abaa168ac fix merge conflict 2024-04-22 07:05:40 -05:00