Vulcan
548bf5df95
Rebase (WIP)
...
Do not clone
2023-08-21 23:53:31 +05:30
Oleg Zabluda
c2834c8a1f
Remove unneeded check of free(NULL)
...
Passing NULL to free() is totally allowed
2023-08-21 10:54:53 -07:00
Andrej
ee95b1bf29
Merge pull request #315 from davidar/vocab_source
...
Fix vocab_source in sample.py
2023-08-21 08:26:28 -07:00
Andrej Karpathy
d02e0c90d8
Merge branch 'rdentato-patch-check-params'
2023-08-21 15:17:37 +00:00
Andrej Karpathy
33d94f60a5
parameter validation cleanup
2023-08-21 15:17:14 +00:00
Remo Dentato
2d972f1763
Merge branch 'karpathy:master' into patch-check-params
2023-08-21 17:02:42 +02:00
Andrej
8a3ea7b433
Merge pull request #329 from atamurad/import_meta
...
Moved export_meta_llama_bin.py to new export.py
2023-08-21 07:34:32 -07:00
atamyrat
61c26d5392
Updated README to replace export_meta_llama_bin.py script with export.py
2023-08-21 14:24:01 +03:00
atamyrat
36a78af5e1
tested load_meta_model() in export.py, deleting old export_meta_llama_bin.py file
2023-08-21 14:19:56 +03:00
atamyrat
de005474d3
Added load_meta_model() to export.py
2023-08-21 14:13:47 +03:00
Vulcan
410e17ea67
Merge remote-tracking branch 'upstream/master'
2023-08-21 16:37:10 +05:30
Vulcan
fc2e5aebd4
Update run.c for manual rebase (WIP)
...
Do not clone (WIP)
2023-08-21 16:36:31 +05:30
Vulcan
2e62eacfbc
Baremetal/Portable Embedding (WIP)
...
Embedding model, tokenizer & assets for baremetal and portable builds (WIP)
2023-08-21 15:18:33 +05:30
rdentato
4444575c4e
Added check of generation parameters.
2023-08-21 06:43:39 +00:00
Andrej Karpathy
dd61b13e57
delete the save_torchscript export file, but copy its content to the new export.py for the future maybe
2023-08-21 05:09:06 +00:00
Andrej Karpathy
ea44f53568
now that the export.py HF functionality is in master, we can delete this file, and update the readme
2023-08-21 04:58:19 +00:00
Andrej
801c68f5a1
Merge pull request #326 from atamurad/import_hf
...
Added huggingface model loader/importer to export.py
2023-08-20 21:53:17 -07:00
Andrej
74a68eeb35
Merge pull request #325 from HarryGifford/users/hegi/update-readme-threading
...
Update readme with suggestion on number of threads to use
2023-08-20 21:50:26 -07:00
Andrej Karpathy
288b3cec09
remove dagger in the eyeball
2023-08-21 04:47:49 +00:00
Andrej Karpathy
14275bd623
minor clean. i think a lot of chaos has been reduced for today. we shall now rest.
2023-08-21 04:43:24 +00:00
Andrej Karpathy
3868f732a4
and finally refactor the Sampler. things are starting to look a lot cleaner I think
2023-08-21 04:23:02 +00:00
Andrej Karpathy
8a377a1d31
refactor the Transformer (Config, Weights, RunState) into a single object, with build and free too
2023-08-21 03:55:12 +00:00
Andrej Karpathy
ae2e4f8d88
name the tokenizer methods cleaner: encode and decode
2023-08-21 03:11:54 +00:00
atamyrat
0dd82158f6
removed transformers from requirements.txt, added error message
2023-08-21 06:07:29 +03:00
atamyrat
155475a523
Fix WQ and WK permutation in huggingface models
2023-08-21 05:16:11 +03:00
atamyrat
d7704bdeaa
mark ModelArgs.hidden_dim as optional and calculate as previously if not provided
2023-08-21 03:40:34 +03:00
atamyrat
09db52c69e
Added huggingface model loader to export.py
2023-08-21 02:59:12 +03:00
Harry Gifford
a72b3b0206
Update readme with suggestion on number of threads to use
...
Update the documentation to make suggestions on the number of threads. The performance difference can be very large. Also linked to the PyTorch docs which are relevant here.
2023-08-20 15:01:33 -07:00
Andrej Karpathy
c74456f3f0
refactor step 1. the tokenizer, and all the other abstractions, are a total mess, refactoring things a bit
2023-08-20 18:18:23 +00:00
Andrej Karpathy
1e335a41cf
remove freq_cis fields as they are not used anymore
2023-08-20 17:26:43 +00:00
Andrej Karpathy
c0511de617
probindex should never have been part of RunState. i apologize for this failure of abstraction
2023-08-20 17:18:06 +00:00
Andrej
8c93c7a30e
Merge pull request #322 from karpathy/feature/export
...
New model export (the code remains "dead" and legacy version is still the default behavior, so no breaking changes are introduced). The major benefit is a new export.py file, which we can use to centralize work on formatting: both imports and exports.
2023-08-20 10:08:32 -07:00
Andrej Karpathy
13dcee493a
todos update
2023-08-20 17:02:22 +00:00
Andrej Karpathy
f3db92a2dc
use out_file.tell() instead of nbytes += arithmetic
2023-08-20 16:51:35 +00:00
Vulcan
f7a7ed94c8
Merge remote-tracking branch 'upstream/master'
2023-08-20 17:19:07 +05:30
Vulcan
bbb7b0e365
Update README.md
2023-08-20 14:37:23 +05:30
Andrej Karpathy
fa8dfd854e
isolate read_checkpoint, because i'd like to now make it support both version 0 and version 1
2023-08-19 19:21:12 +00:00
Andrej Karpathy
4df5e2e939
make version 1 be the legacy export but with new header. version 2 will be Q8_0 export
2023-08-19 18:51:32 +00:00
Andrej Karpathy
4212bd6d43
oops fix double indent on quantize def
2023-08-19 18:34:49 +00:00
Andrej Karpathy
7f551dbfd7
new model export: versions 0 (legacy) and 1
2023-08-19 18:25:20 +00:00
Andrej
6c5d78fa41
Merge pull request #317 from yiminghan/yhan/old
...
Add a link to Dart port in README
2023-08-19 10:01:08 -07:00
Andrej
db1a722816
Merge pull request #318 from rahoua/master
...
YARP - Yet Another Rust Port in README.md
2023-08-19 10:00:56 -07:00
Andrej
d2a546c577
Merge pull request #319 from RahulSChand/warning
...
Give better error message in Tinystories data loader
2023-08-19 10:00:27 -07:00
rahulschand
fbefeec1b1
add assert message to give better warning
2023-08-19 13:05:26 +05:30
rahoua
978c311b30
Add pecca-rs to README.md
2023-08-18 14:58:21 -07:00
YiMing Han
882e480bc0
update read me
2023-08-18 15:18:29 -04:00
YiMing Han
d09ebbb32b
Revert "working one"
...
This reverts commit 8607b11ea1
2023-08-18 15:14:08 -04:00
YiMing Han
bc7cb7d0e8
Revert "only dart"
...
This reverts commit 01df3731d6
2023-08-18 15:13:59 -04:00
YiMing Han
01df3731d6
only dart
2023-08-18 15:09:24 -04:00
YiMing Han
8607b11ea1
working one
2023-08-18 15:07:41 -04:00
