chenyu f7f67e0cc5 simple fix llama shard with quantize (#3882) ...

copy scale on all device for now. naive sharding does not work because scale needs expand to really save memory.

70B does not work due to HSA_STATUS_ERROR_OUT_OF_RESOURCES.

`python3 examples/llama.py --gen 2 --size 13B --shard 6 --prompt "Hello." --count 10 --temperature 0 --timing --quantize`

13B on 6 gpus uses 47 GB v.s. 34 GB quantized

2024-03-22 18:15:37 -04:00

..

conversation_data

Whisper + LLAMA + VITS (#2332)

2023-12-02 15:03:46 -08:00

mlperf

estimated resnet training time for BENCHMARK (#3769)

2024-03-15 22:36:58 -04:00

rl

more beautiful_cartpole with exposed hparams

2024-01-07 17:41:09 -08:00

sovits_helpers

move dtypes to dtype.py (#2964)

2024-01-01 14:58:48 -08:00

vgg7_helpers

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00

webgl/yolov8

webgl backend in extra (#3041)

2024-01-08 09:29:13 -08:00

webgpu/stable_diffusion

s/lazydata.realized/lazydata.base.realized/g (#2914)

2023-12-22 14:45:13 -05:00

__init__.py

failing llama test

2023-03-11 16:28:10 -08:00

beautiful_cartpole.py

more beautiful_cartpole with exposed hparams

2024-01-07 17:41:09 -08:00

beautiful_mnist.py

move sample inside jit for beautiful_mnist (#3115)

2024-01-14 01:36:30 -05:00

benchmark_train_efficientnet.py

move graph.py and jit.py into features (#3376)

2024-02-12 17:34:34 +01:00

coder.py

apply the same fix_bf16 in llama and coder (#3789)

2024-03-17 21:25:24 -04:00

compile_efficientnet.py

webgl backend in extra (#3041)

2024-01-08 09:29:13 -08:00

compile_tensorflow.py

updating to work with new internal apis (#2755)

2023-12-13 21:54:47 -08:00

conversation.py

fix conversation: llama generates token not prob now (#3120)

2024-01-14 13:10:01 -05:00

efficientnet.py

move graph.py and jit.py into features (#3376)

2024-02-12 17:34:34 +01:00

f16_w_uint32.py

move dtypes to dtype.py (#2964)

2024-01-01 14:58:48 -08:00

gpt2.py

simple LoadOps.ASSIGN (#3745)

2024-03-14 20:44:34 -07:00

handcode_resnet50_opt.py

fixing the benchmark not printing in handcode resnet50 opt example (#3850)

2024-03-21 00:55:31 -04:00

hlb_cifar10.py

search: change to use "spawn" and limit the number of tasks per child (#3862)

2024-03-21 21:23:36 -07:00

index.html

Enable Multi-Output Export (#2179)

2023-10-30 18:42:26 -07:00

llama.py

simple fix llama shard with quantize (#3882)

2024-03-22 18:15:37 -04:00

mask_rcnn.py

move things, clean up extra (#2292)

2023-11-13 20:18:40 -08:00

mixtral.py

remove mixtral weight to clang first (#3792)

2024-03-17 23:33:17 -04:00

mnist_gan.py

move state to nn/state (#1619)

2023-08-22 07:36:24 -07:00

serious_mnist.py

move state to nn/state (#1619)

2023-08-22 07:36:24 -07:00

simple_conv_bn.py

with Tensor.train() (#1935)

2023-09-28 18:02:31 -07:00

so_vits_svc.py

move dtypes to dtype.py (#2964)

2024-01-01 14:58:48 -08:00

stable_diffusion_seed0.png

validate stable diffusion for seed 0 (#2773)

2023-12-15 00:07:09 -05:00

stable_diffusion.py

add Tensor.replace (#3738)

2024-03-14 13:34:14 -07:00

train_efficientnet.py

move things, clean up extra (#2292)

2023-11-13 20:18:40 -08:00

train_resnet.py

move things, clean up extra (#2292)

2023-11-13 20:18:40 -08:00

transformer.py

fix onehot and jit in examples/transformer (#3073)

2024-01-10 02:22:41 -05:00

vgg7.py

waifu2x vgg7: testcase, auto-RGBA->RGB, function to grab pretrained models, training "fix" (#2117)

2023-10-19 22:07:15 -07:00

vit.py

move to new cached fetch (#2493)

2023-11-28 17:36:55 -08:00

vits.py

fix example that calls Tensor.__bool__ (#3650)

2024-03-07 16:59:26 -05:00

whisper.py

move graph.py and jit.py into features (#3376)

2024-02-12 17:34:34 +01:00

yolov3.py

Update yolov3.py (#2680)

2023-12-08 12:59:38 -08:00

yolov8-onnx.py

[ready] Replacing os with pathlib (#1708)

2023-08-30 10:41:08 -07:00

yolov8.py

fix yolov8.py (#3742)

2024-03-14 17:33:45 -04:00