Instructions to use xiaomoguhzz/VisionEncoder with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use xiaomoguhzz/VisionEncoder with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("xiaomoguhzz/VisionEncoder", device_map="auto") - Notebooks
- Google Colab
- Kaggle
add V10.x cache vjepa21/video_10pct_chunks
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/dataset_info.json +60 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/state.json +1 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/vjepa21_patch_meta.json +12 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_12/train/data-00000-of-00001.arrow +3 -0
- data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_12/train/dataset_info.json +60 -0
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b795496c9c302e4d189d3d026bf1e414700039878502561378ddf58362a7ab83
|
| 3 |
+
size 7837872
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7447721,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_00.json": {
|
| 10 |
+
"num_bytes": 8441128,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8441128,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15888849,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7447721,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "fcdb0d51c472f4c5", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_00/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_00/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3414.91,
|
| 10 |
+
"lengths_sample_mean_new": 4886.91,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:42"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0631e215c50f00f42c5c05c315bca64a097eb1ba41b7eb9ec83ddd1ca1a9409
|
| 3 |
+
size 7760272
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7368826,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_01.json": {
|
| 10 |
+
"num_bytes": 8374839,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8374839,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15743665,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7368826,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "1fa6968a695d7ed9", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_01/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_01/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3423.39,
|
| 10 |
+
"lengths_sample_mean_new": 4895.39,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:54"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2543b8b3429a0b696c74b11b53393d337f8682179423d3b2013708d59a0ab0d
|
| 3 |
+
size 7853432
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7463924,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_02.json": {
|
| 10 |
+
"num_bytes": 8451092,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8451092,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15915016,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7463924,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "67922778fe789d97", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_02/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_02/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3379.45,
|
| 10 |
+
"lengths_sample_mean_new": 4851.45,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:55"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:492dfdbc3d3fa27eab18aac4ba67a2deb3bf1d3b911d27885de11b3bb6bccafa
|
| 3 |
+
size 7934096
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7544302,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_03.json": {
|
| 10 |
+
"num_bytes": 8531728,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8531728,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 16076030,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7544302,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "034aaf22051fa0ae", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_03/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_03/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3382.52,
|
| 10 |
+
"lengths_sample_mean_new": 4854.52,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:56"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c22a22113c5e2e046241aff63714ddcd44cb8c8140e553aa9df7fc3f66128210
|
| 3 |
+
size 7676528
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7285950,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_04.json": {
|
| 10 |
+
"num_bytes": 8279218,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8279218,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15565168,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7285950,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "5fa0e5d6216678a0", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_04/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_04/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3430.91,
|
| 10 |
+
"lengths_sample_mean_new": 4902.91,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:56"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:762b517a0a1e6a9b326029cfd36d2887cccf8d0fe5ccd8764e902a33bcc898d4
|
| 3 |
+
size 7741288
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7351520,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_05.json": {
|
| 10 |
+
"num_bytes": 8342040,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8342040,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15693560,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7351520,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "b220e4881c83491d", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_05/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_05/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3404.79,
|
| 10 |
+
"lengths_sample_mean_new": 4876.79,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:57"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47774d9c292ae6dd9207825f1449ac86267dafbcac1ad6534c05321c67af6f59
|
| 3 |
+
size 7977632
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7588693,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_06.json": {
|
| 10 |
+
"num_bytes": 8571022,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8571022,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 16159715,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7588693,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "699bcf19aee2da88", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_06/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_06/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3432.08,
|
| 10 |
+
"lengths_sample_mean_new": 4904.08,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:58"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab4612f3e5fb5d040685446c5a7b01a3bc0793307086fc6b7aff195914e3ce16
|
| 3 |
+
size 7979912
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7590275,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_07.json": {
|
| 10 |
+
"num_bytes": 8581233,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8581233,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 16171508,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7590275,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "f15df0b8a41351d5", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_07/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_07/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3467.97,
|
| 10 |
+
"lengths_sample_mean_new": 4939.97,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:58"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50747840a95d741f81227cca68f4e2e3ea23d2b3753f9d2a0b0da046a5f9de02
|
| 3 |
+
size 7878016
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7487777,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_08.json": {
|
| 10 |
+
"num_bytes": 8478861,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8478861,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15966638,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7487777,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "efb20606a7897f10", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_08/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_08/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3423.25,
|
| 10 |
+
"lengths_sample_mean_new": 4895.25,
|
| 11 |
+
"timestamp": "2026-05-11T15:38:59"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:087cd98bc9fe8f0a0ec829c69a6cb3858308f95539634c3962e87d837a468ef3
|
| 3 |
+
size 7931944
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7541773,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_09.json": {
|
| 10 |
+
"num_bytes": 8536644,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8536644,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 16078417,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7541773,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "80f39657123e1abf", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_09/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_09/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3423.52,
|
| 10 |
+
"lengths_sample_mean_new": 4895.52,
|
| 11 |
+
"timestamp": "2026-05-11T15:39:00"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84196cb29e4f2b038df58d8b36196f09e2479b30da98901093e277dfbc766906
|
| 3 |
+
size 8073904
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7684339,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_10.json": {
|
| 10 |
+
"num_bytes": 8671454,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8671454,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 16355793,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7684339,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "40d0c61caab0704d", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_10/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_10/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3450.13,
|
| 10 |
+
"lengths_sample_mean_new": 4922.13,
|
| 11 |
+
"timestamp": "2026-05-11T15:39:00"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8c9d335089cc30d89c1fadc34fa7ef9ec4892d7995bde9cb3e0186b9449b14e
|
| 3 |
+
size 7786696
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7397014,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_11.json": {
|
| 10 |
+
"num_bytes": 8385641,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8385641,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15782655,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7397014,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/state.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"_data_files": [{"filename": "data-00000-of-00001.arrow"}], "_fingerprint": "f24435657501f2f2", "_format_columns": null, "_format_kwargs": {}, "_format_type": null, "_output_all_columns": false, "_split": null}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_11/train/vjepa21_patch_meta.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"patcher": "siglip2_to_vjepa21",
|
| 3 |
+
"mode": "video",
|
| 4 |
+
"from": "siglip2 (SigLIP2 base 16f \u00d7 196 + 1 = 3137 video tokens)",
|
| 5 |
+
"to": "vjepa21 (V-JEPA 2.1 16f \u00d7 288 + 1 = 4609 video tokens)",
|
| 6 |
+
"delta_formula": "lengths[0] += 1472 * len(videos)",
|
| 7 |
+
"input": "/share/m2v_intern_v3/wangjunjie09/VisionEncoder/data/vmllm_cached/siglip2/video_10pct_chunks/chunk_11/train",
|
| 8 |
+
"rows": 5681,
|
| 9 |
+
"lengths_sample_mean_old": 3410.52,
|
| 10 |
+
"lengths_sample_mean_new": 4882.52,
|
| 11 |
+
"timestamp": "2026-05-11T15:39:01"
|
| 12 |
+
}
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_12/train/data-00000-of-00001.arrow
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f4b94e087f25942e59e0fe5bb89a165dc4895d3f0ba80c8c8c4988828e35b7c
|
| 3 |
+
size 7788360
|
data/vmllm_cached/vjepa21/video_10pct_chunks/chunk_12/train/dataset_info.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"builder_name": "json",
|
| 3 |
+
"citation": "",
|
| 4 |
+
"config_name": "default",
|
| 5 |
+
"dataset_name": "json",
|
| 6 |
+
"dataset_size": 7397632,
|
| 7 |
+
"description": "",
|
| 8 |
+
"download_checksums": {
|
| 9 |
+
"/share/m2v_intern_v3/wangjunjie09/model_cache/VisionEncoder/ms-swift-data/video_sft_chunks_local/chunk_12.json": {
|
| 10 |
+
"num_bytes": 8394845,
|
| 11 |
+
"checksum": null
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"download_size": 8394845,
|
| 15 |
+
"features": {
|
| 16 |
+
"messages": [
|
| 17 |
+
{
|
| 18 |
+
"content": {
|
| 19 |
+
"dtype": "string",
|
| 20 |
+
"_type": "Value"
|
| 21 |
+
},
|
| 22 |
+
"role": {
|
| 23 |
+
"dtype": "string",
|
| 24 |
+
"_type": "Value"
|
| 25 |
+
}
|
| 26 |
+
}
|
| 27 |
+
],
|
| 28 |
+
"videos": {
|
| 29 |
+
"feature": {
|
| 30 |
+
"dtype": "string",
|
| 31 |
+
"_type": "Value"
|
| 32 |
+
},
|
| 33 |
+
"_type": "Sequence"
|
| 34 |
+
},
|
| 35 |
+
"lengths": {
|
| 36 |
+
"feature": {
|
| 37 |
+
"dtype": "int64",
|
| 38 |
+
"_type": "Value"
|
| 39 |
+
},
|
| 40 |
+
"_type": "Sequence"
|
| 41 |
+
}
|
| 42 |
+
},
|
| 43 |
+
"homepage": "",
|
| 44 |
+
"license": "",
|
| 45 |
+
"size_in_bytes": 15792477,
|
| 46 |
+
"splits": {
|
| 47 |
+
"train": {
|
| 48 |
+
"name": "train",
|
| 49 |
+
"num_bytes": 7397632,
|
| 50 |
+
"num_examples": 5681,
|
| 51 |
+
"dataset_name": "json"
|
| 52 |
+
}
|
| 53 |
+
},
|
| 54 |
+
"version": {
|
| 55 |
+
"version_str": "0.0.0",
|
| 56 |
+
"major": 0,
|
| 57 |
+
"minor": 0,
|
| 58 |
+
"patch": 0
|
| 59 |
+
}
|
| 60 |
+
}
|