Instructions to use lingkai/open-clip with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- OpenCLIP
How to use lingkai/open-clip with OpenCLIP:
import open_clip model, preprocess_train, preprocess_val = open_clip.create_model_and_transforms('hf-hub:lingkai/open-clip') tokenizer = open_clip.get_tokenizer('hf-hub:lingkai/open-clip') - Notebooks
- Google Colab
- Kaggle
Upload folder using huggingface_hub
Browse files- checkpoints/epoch_25.pt +3 -0
- checkpoints/epoch_26.pt +3 -0
- checkpoints/epoch_27.pt +3 -0
- checkpoints/epoch_28.pt +3 -0
- checkpoints/epoch_29.pt +3 -0
- checkpoints/epoch_30.pt +3 -0
- checkpoints/epoch_31.pt +3 -0
- checkpoints/epoch_32.pt +3 -0
- out.log +0 -0
- params.txt +128 -0
checkpoints/epoch_25.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:44fd948ffd4b4e8b4f0ed47d0f63cbb71cb796782ff36043cad17273b0a97e5b
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_26.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b908c7786324cba405ca314b16303145f1d73d0c7c88fb8258e5b52aa1a9b29
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_27.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8c5391b8c0b076af12205680f56f75eb38f02b1e503509e4813496d6237ce619
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_28.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:063b8dea65a2dd450e4ce80801d2bee99fe95a847bc5c03af8d38e7799d9fd98
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_29.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9511503638aece912c7ea4de3eac2a6a20c14c15a94363e8501c60d8ce922035
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_30.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3aefc59072fa830757c9d6245583cb6aceede9fd47a1dca4bfa497206293236a
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_31.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c86249976545854448be0ffd1c4cf5b8e2d0a7a0118b936067b62a61a2c4e7c6
|
| 3 |
+
size 1815707357
|
checkpoints/epoch_32.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccc19a20a89a66811d6fad83d2fee21a3287abf4be87a56067721b5dc2dfac88
|
| 3 |
+
size 1815707357
|
out.log
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
params.txt
ADDED
|
@@ -0,0 +1,128 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
accum_freq: 1
|
| 2 |
+
audio_ext: flac
|
| 3 |
+
audio_fill: repeatpad
|
| 4 |
+
audio_fusion: False
|
| 5 |
+
audio_int16_normalize: False
|
| 6 |
+
audio_trunc: rand_trunc
|
| 7 |
+
audio_zeroshot_audio_key: audio
|
| 8 |
+
audio_zeroshot_class_key: category
|
| 9 |
+
audio_zeroshot_dataset: None
|
| 10 |
+
audio_zeroshot_multiprocessing_context: forkserver
|
| 11 |
+
audio_zeroshot_split: train
|
| 12 |
+
audio_zeroshot_target_key: target
|
| 13 |
+
audio_zeroshot_templates: None
|
| 14 |
+
audio_zeroshot_workers: 0
|
| 15 |
+
aug_cfg: {}
|
| 16 |
+
batch_size: 4096
|
| 17 |
+
beta1: 0.9
|
| 18 |
+
beta2: 0.98
|
| 19 |
+
cache_dir: None
|
| 20 |
+
checkpoint_path: ./logs/ViT-B-32-Vanilla-resume/checkpoints
|
| 21 |
+
coca_caption_loss_weight: 2.0
|
| 22 |
+
coca_contrastive_loss_weight: 1.0
|
| 23 |
+
copy_codebase: False
|
| 24 |
+
csv_caption_key: title
|
| 25 |
+
csv_img_key: filepath
|
| 26 |
+
csv_separator:
|
| 27 |
+
dataset_resampled: False
|
| 28 |
+
dataset_type: webdataset
|
| 29 |
+
ddp_static_graph: False
|
| 30 |
+
debug: False
|
| 31 |
+
delete_previous_checkpoint: False
|
| 32 |
+
device: cuda:0
|
| 33 |
+
dist_backend: None
|
| 34 |
+
dist_url: None
|
| 35 |
+
distill: False
|
| 36 |
+
distill_model: None
|
| 37 |
+
distill_pretrained: None
|
| 38 |
+
distributed: True
|
| 39 |
+
epochs: 32
|
| 40 |
+
epochs_cooldown: None
|
| 41 |
+
eps: 1e-06
|
| 42 |
+
force_context_length: None
|
| 43 |
+
force_custom_text: False
|
| 44 |
+
force_image_size: None
|
| 45 |
+
force_naflex_vision: False
|
| 46 |
+
force_patch_dropout: None
|
| 47 |
+
force_quick_gelu: False
|
| 48 |
+
fsdp: False
|
| 49 |
+
fsdp_checkpoint: full
|
| 50 |
+
fsdp_no_reshard_after_forward: False
|
| 51 |
+
fsdp_offload_cpu: False
|
| 52 |
+
gather_with_grad: True
|
| 53 |
+
grad_checkpointing: False
|
| 54 |
+
grad_clip_norm: None
|
| 55 |
+
image_interpolation: None
|
| 56 |
+
image_mean: None
|
| 57 |
+
image_resize_mode: None
|
| 58 |
+
image_std: None
|
| 59 |
+
imagenet_v2: None
|
| 60 |
+
imagenet_val: None
|
| 61 |
+
local_loss: True
|
| 62 |
+
local_rank: 0
|
| 63 |
+
lock_image: False
|
| 64 |
+
lock_image_freeze_bn_stats: False
|
| 65 |
+
lock_image_unlocked_groups: 0
|
| 66 |
+
lock_text: False
|
| 67 |
+
lock_text_freeze_layer_norm: False
|
| 68 |
+
lock_text_unlocked_layers: 0
|
| 69 |
+
log_every_n_steps: 100
|
| 70 |
+
log_level: 20
|
| 71 |
+
log_local: False
|
| 72 |
+
log_path: ./logs/ViT-B-32-Vanilla-resume/out.log
|
| 73 |
+
logs: ./logs/
|
| 74 |
+
loss_dist_impl: None
|
| 75 |
+
lr: 0.0005
|
| 76 |
+
lr_cooldown_end: 0.0
|
| 77 |
+
lr_cooldown_power: 1.0
|
| 78 |
+
lr_scheduler: cosine
|
| 79 |
+
model: ViT-B-32
|
| 80 |
+
momentum: None
|
| 81 |
+
naflex_batch_divisor: 8
|
| 82 |
+
naflex_loss_scale: none
|
| 83 |
+
naflex_max_image_tokens_per_batch: 16384
|
| 84 |
+
naflex_num_train_image_tokens: None
|
| 85 |
+
naflex_patch_size_probs: None
|
| 86 |
+
naflex_patch_sizes: None
|
| 87 |
+
naflex_seq_lens: None
|
| 88 |
+
name: ViT-B-32-Vanilla-resume
|
| 89 |
+
no_set_device_rank: False
|
| 90 |
+
opt: adamw
|
| 91 |
+
precision: amp_bfloat16
|
| 92 |
+
pretrained:
|
| 93 |
+
pretrained_audio: None
|
| 94 |
+
pretrained_image: False
|
| 95 |
+
rank: 0
|
| 96 |
+
remote_sync: None
|
| 97 |
+
remote_sync_frequency: 300
|
| 98 |
+
remote_sync_protocol: s3
|
| 99 |
+
report_to: wandb
|
| 100 |
+
resume: /scratch/work/zhul2/code/open_clip/logs/ViT-B-32-Vanilla/checkpoints/epoch_24.pt
|
| 101 |
+
save_frequency: 1
|
| 102 |
+
save_most_recent: False
|
| 103 |
+
seed: 0
|
| 104 |
+
siglip: False
|
| 105 |
+
skip_scheduler: False
|
| 106 |
+
tensorboard: False
|
| 107 |
+
tensorboard_path:
|
| 108 |
+
torchcompile: True
|
| 109 |
+
torchcompile_backend: None
|
| 110 |
+
torchcompile_mode: None
|
| 111 |
+
torchcompile_strategy: task
|
| 112 |
+
train_data: /scratch/shareddata/dldata/laion400M/img2dataset/laion400m-data/{00000..41407}.tar
|
| 113 |
+
train_data_upsampling_factors: None
|
| 114 |
+
train_num_samples: 268836185
|
| 115 |
+
use_bn_sync: False
|
| 116 |
+
use_bnb_linear: None
|
| 117 |
+
use_naflex: False
|
| 118 |
+
val_data: None
|
| 119 |
+
val_frequency: 1
|
| 120 |
+
val_num_samples: None
|
| 121 |
+
wandb: True
|
| 122 |
+
wandb_notes:
|
| 123 |
+
wandb_project_name: open-clip
|
| 124 |
+
warmup: 2000
|
| 125 |
+
wd: 0.2
|
| 126 |
+
workers: 12
|
| 127 |
+
world_size: 8
|
| 128 |
+
zeroshot_frequency: 2
|