Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- V4.3-ckpt/ckpt_iter0005000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0005000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0010000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0010000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0015000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0015000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0020000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0020000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0025000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0025000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0030000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0030000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0035000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0035000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0040000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0040000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0045000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0045000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0050000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0050000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0055000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0055000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0060000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0060000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0065000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0065000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0070000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0070000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0075000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0075000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0080000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0080000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0085000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0085000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0090000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0090000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0095000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- V4.3-ckpt/ckpt_iter0095000.pth/mp_rank_00_model_states.pt +3 -0
- V4.3-ckpt/config.json +243 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0000000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0001000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0002000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0003000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0004000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0005000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0006000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0007000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0008000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0009000.png +0 -0
- V4.3-ckpt/cosine_sim_graphs/graph_step_0010000.png +0 -0
V4.3-ckpt/ckpt_iter0005000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:56c056016d8b8a248e7d54c844b6c004cfe54e1e483921f5b1952b065ca85349
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0005000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b531154b45173908020ee9a162a8100f9cc3d8d10deb37b09d68389308a6135d
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0010000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:da64c8d2b0268d31f45c524311146f16e9cbb688d917100e125834b617c2c9b5
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0010000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b380f89583e0e2a2ca080f5ba4036d78b87c669669dd7a6fb2d5ba9d3cda83dd
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0015000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d81bfb74f3b4f2052f26112507bdfae7927aa34eb499538107394f7a960d3c8c
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0015000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7f49055976a46362bf30c1be8ca9157e66a90f2475ed9c76cf0f8137cd86f05b
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0020000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e75ff5e61aa6a945214debc3e1a795d94942ff9a808dc0f6477e822cad626c4b
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0020000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:62837c90687bd793e4f6e7d8913c6f6d151f4d61ddacbcb69d10c29a20fe86a7
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0025000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7ca26f145d58915af43963f5c3801eedccbbf4206a50fa85cab73d7835811a5d
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0025000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c323dd0af112f1acb8c4b73beda2e326aa6f5e78d3c77e65b69aab953e49ead0
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0030000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bee656e4312a9700e9e379b217f3e550c356f2ccda11b15e01c4c3f0e22da5ae
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0030000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5d7b9ee7f0523c215814a1ca476a77847a79eff99d1b39acb3a15287ba0e5d27
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0035000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8da5825bf83934e65a7210eb7cd1e0bc057bb3a1d1554c9238f559cb8314a361
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0035000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9ea7171231f41b203c368337ff928a17e16c81df3e0d843a1915594d8e26f495
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0040000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c5448f30ccdae36c771d6945b984aff7383872024ea17b1afc659f6a8dea2fb6
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0040000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e70668941469527626a4fa068217258aa7d5269f56d5beaa4da58395e5928ab5
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0045000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:046269135cd8be837f2397b20a241cd657b3ed8146689ebc03f31d416e96aeb4
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0045000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7eb4432e39cf49637932611eb23f53f5259e0e97936500334183a0d1f872eaa6
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0050000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:39a023ee1da3bdee2870c96ecc671bc6a877826f3d78939e79fd6705d7133439
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0050000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5d9e063bde8ee672fcad75c4f9618615f6edb7c18350b56c04973a2723f6c45
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0055000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f1da6e6b735e6d45595a608e43ae7d161deb7922f02757636e5b0d3ee06e50e7
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0055000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e58283a54fa4c0ddb9d757b5e7f40dd4fcd4a4a0fa1b40ae5a7b69c5354b5741
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0060000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:03d2ad0d1fc3ac4f3d0dbf083955bbfaf32b893c1221b80cb4f7b430bbf3ea45
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0060000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8bc97f15576086587cb20c2745cf41c59881c3e2d3d9df2915f8d157ca7d815a
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0065000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8aead2cd83a525d092ebd67b163a32c1b26f115f5d990a63d5d89d8bc7354eef
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0065000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9a971fc817b840f3f5f0dce4dbdafec4342e4e6879bf1ac464cade9ad033db11
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0070000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dd8c3d952773317fc8c7e86456c26d63cb3a0c97d7caaa2c6fd62ec7e39135dd
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0070000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ff32e409e5a8d8b3f7f71edd2e8cb251890ea7a2be606c6ca03c6a931673c10a
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0075000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3ed28a79904a7820cbf7759416ff99f0bd13fa938f88192e05b637f641297ad3
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0075000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aa06fb8ebce63e2be53630a27cca6e736e2d62691c409cab3f8bba9768fb8c69
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0080000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:360fa7b5e63ace9b88522feea8df6c4e3ce1c4d2d40fd0da960d582e6284af4b
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0080000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e33aed92487ae27d3fa58c8f3d986c4c9bfa4e5cb8c6ffa0023047b8912b0441
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0085000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bae8dd84fdfae832e273b3f4495533d441105ad22a23f09a7721f1fdac88c4b9
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0085000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:320ed1fa617cfd7e59180f4035c35a0ba9483500e8747832c532b2304e9fcd26
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0090000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b5356ed18783ee39e805ebb55315be818e386adfbd5b4af38c4d3c41bf9696c0
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0090000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f66984aab3c7fdf34c7058e19c6a175e1d93838e5e7c39cbc4702a221ef1b384
|
3 |
+
size 26443493
|
V4.3-ckpt/ckpt_iter0095000.pth/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:878e1ccefc862a1c2263801b39bac100dfb3586596988f10504cd45e69c7bc6f
|
3 |
+
size 158369093
|
V4.3-ckpt/ckpt_iter0095000.pth/mp_rank_00_model_states.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1e6979b3666f1766dc13530a5eea3bd87d317550fd3849b8b3de9c5a9c3e0858
|
3 |
+
size 26443493
|
V4.3-ckpt/config.json
ADDED
@@ -0,0 +1,243 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"root_path": "/home/zli",
|
3 |
+
"available_corpus": {
|
4 |
+
"cc3m": {
|
5 |
+
"anno_path": "your_path",
|
6 |
+
"data_root": "",
|
7 |
+
"media_type": "image"
|
8 |
+
},
|
9 |
+
"webvid_10m": {
|
10 |
+
"anno_path": "your_path",
|
11 |
+
"data_root": "",
|
12 |
+
"media_type": "video"
|
13 |
+
},
|
14 |
+
"smol_test": {
|
15 |
+
"anno_path": "/root/IV2/InternVideo2/multi_modality/data_test/smol_test.json",
|
16 |
+
"data_root": "/root/IV2/InternVideo2/multi_modality/data_test/",
|
17 |
+
"media_type": "video"
|
18 |
+
},
|
19 |
+
"slim_kinetics": {
|
20 |
+
"anno_path": "/home/zli/kinetics-dataset/k600/train/train/kinetics_v2.json",
|
21 |
+
"data_root": "/home/zli/kinetics-dataset/k600/train/train",
|
22 |
+
"media_type": "video"
|
23 |
+
},
|
24 |
+
"slim_kinetics_act_val": {
|
25 |
+
"anno_path": "/home/zli/kinetics-dataset/k600/test/kinetics-test.json",
|
26 |
+
"data_root": "/home/zli/kinetics-dataset/k600/test/",
|
27 |
+
"media_type": "video",
|
28 |
+
"is_act_rec": true
|
29 |
+
}
|
30 |
+
},
|
31 |
+
"VisionEncoders": {},
|
32 |
+
"TextEncoders": {
|
33 |
+
"bert": {
|
34 |
+
"name": "bert_base",
|
35 |
+
"pretrained": "bert-base-uncased",
|
36 |
+
"config": "configs/config_bert.json",
|
37 |
+
"d_model": 768,
|
38 |
+
"fusion_layer": 9
|
39 |
+
},
|
40 |
+
"bert_large": {
|
41 |
+
"name": "bert_large",
|
42 |
+
"pretrained": "bert-large-uncased",
|
43 |
+
"config": "configs/config_bert_large.json",
|
44 |
+
"d_model": 1024,
|
45 |
+
"fusion_layer": 19
|
46 |
+
},
|
47 |
+
"med_bert": {
|
48 |
+
"name": "med_bert_base",
|
49 |
+
"pretrained": "bert-base-uncased",
|
50 |
+
"config": "configs/med_config.json",
|
51 |
+
"d_model": 768
|
52 |
+
},
|
53 |
+
"med_bert_large": {
|
54 |
+
"name": "med_bert_large",
|
55 |
+
"pretrained": "bert-base-uncased",
|
56 |
+
"config": "configs/med_large_config.json",
|
57 |
+
"d_model": 768
|
58 |
+
}
|
59 |
+
},
|
60 |
+
"train_corpus": "slim_kinetics",
|
61 |
+
"train_file": {
|
62 |
+
"anno_path": "/home/zli/kinetics-dataset/k600/train/train/kinetics_v2.json",
|
63 |
+
"data_root": "/home/zli/kinetics-dataset/k600/train/train",
|
64 |
+
"media_type": "video"
|
65 |
+
},
|
66 |
+
"test_file": {
|
67 |
+
"act_val": {
|
68 |
+
"anno_path": "/home/zli/kinetics-dataset/k600/test/kinetics-test.json",
|
69 |
+
"data_root": "/home/zli/kinetics-dataset/k600/test/",
|
70 |
+
"media_type": "video",
|
71 |
+
"is_act_rec": true
|
72 |
+
}
|
73 |
+
},
|
74 |
+
"test_types": [
|
75 |
+
"act_val"
|
76 |
+
],
|
77 |
+
"num_workers": 2,
|
78 |
+
"stop_key": null,
|
79 |
+
"num_frames": 8,
|
80 |
+
"num_frames_test": 8,
|
81 |
+
"batch_size": 16,
|
82 |
+
"batch_size_test": 16,
|
83 |
+
"max_txt_l": 32,
|
84 |
+
"size_t": 224,
|
85 |
+
"inputs": {
|
86 |
+
"image_res": 224,
|
87 |
+
"video_input": {
|
88 |
+
"num_frames": 8,
|
89 |
+
"sample_type": "all",
|
90 |
+
"num_frames_test": 8,
|
91 |
+
"sample_type_test": "all",
|
92 |
+
"random_aug": false
|
93 |
+
},
|
94 |
+
"max_txt_l": {
|
95 |
+
"image": 32,
|
96 |
+
"video": 32
|
97 |
+
},
|
98 |
+
"batch_size": {
|
99 |
+
"image": 16,
|
100 |
+
"video": 16
|
101 |
+
},
|
102 |
+
"batch_size_test": {
|
103 |
+
"image": 16,
|
104 |
+
"video": 16
|
105 |
+
}
|
106 |
+
},
|
107 |
+
"model": {
|
108 |
+
"model_cls": "InternVideo2_CLIP_small",
|
109 |
+
"vision_encoder": {
|
110 |
+
"name": "internvideo2",
|
111 |
+
"in_chans": 3,
|
112 |
+
"patch_size": 14,
|
113 |
+
"img_size": 224,
|
114 |
+
"qkv_bias": false,
|
115 |
+
"drop_path_rate": 0.0,
|
116 |
+
"head_drop_path_rate": 0.0,
|
117 |
+
"embed_dim": 768,
|
118 |
+
"num_heads": 12,
|
119 |
+
"mlp_ratio": 4,
|
120 |
+
"init_values": 0.1,
|
121 |
+
"qk_normalization": true,
|
122 |
+
"depth": 12,
|
123 |
+
"use_flash_attn": true,
|
124 |
+
"use_fused_rmsnorm": true,
|
125 |
+
"use_fused_mlp": true,
|
126 |
+
"fused_mlp_heuristic": 1,
|
127 |
+
"drop_cls_token": false,
|
128 |
+
"attn_pool_num_heads": 16,
|
129 |
+
"clip_embed_dim": 768,
|
130 |
+
"layerscale_no_force_fp32": true,
|
131 |
+
"num_frames": 8,
|
132 |
+
"tubelet_size": 1,
|
133 |
+
"sep_pos_embed": false,
|
134 |
+
"use_checkpoint": false,
|
135 |
+
"checkpoint_num": 0,
|
136 |
+
"align_dim": 512
|
137 |
+
},
|
138 |
+
"streaming_vision_encoder": {
|
139 |
+
"in_chans": 3,
|
140 |
+
"patch_size": 14,
|
141 |
+
"img_size": 224,
|
142 |
+
"vit_qkv_bias": true,
|
143 |
+
"vit_drop_path_rate": 0.05,
|
144 |
+
"student_embed_dim": 384,
|
145 |
+
"student_depth": 4,
|
146 |
+
"student_num_heads": 6,
|
147 |
+
"vit_mlp_ratio": 3.0,
|
148 |
+
"vit_init_values": null,
|
149 |
+
"vit_qk_normalization": false,
|
150 |
+
"vit_sep_pos_embed": true,
|
151 |
+
"vit_norm_layer_type": "rmsnorm",
|
152 |
+
"rnn_type": "lstm",
|
153 |
+
"rnn_hidden_size": 1024,
|
154 |
+
"rnn_num_layers": 1,
|
155 |
+
"fc_hidden_layers": [],
|
156 |
+
"teacher_clip_embed_dim": 768,
|
157 |
+
"student_num_frames_processed_by_vit": 1,
|
158 |
+
"student_tubelet_size_for_vit": 1
|
159 |
+
},
|
160 |
+
"text_encoder": {
|
161 |
+
"name": "mobileclip_b"
|
162 |
+
},
|
163 |
+
"temp": 0.01,
|
164 |
+
"temp_min": 0.01,
|
165 |
+
"freeze_vision": true,
|
166 |
+
"open_vision_clip_projector": false,
|
167 |
+
"freeze_text": true,
|
168 |
+
"open_text_projection": false,
|
169 |
+
"open_text_lora": false,
|
170 |
+
"vision_ckpt_path": "/home/zli/IV2/models/stage1/B14/B14_dist_1B_stage2/pytorch_model.bin",
|
171 |
+
"load_vision_ckpt_from_internvideo2_stage2": false,
|
172 |
+
"text_ckpt_path": "/home/zli/IV2/models/mobileclip_blt.pt",
|
173 |
+
"extra_ckpt_path": "/home/zli/IV2/models/clip/B14/pytorch_model.bin"
|
174 |
+
},
|
175 |
+
"criterion": {
|
176 |
+
"loss_weight": {
|
177 |
+
"vtc": 1.0
|
178 |
+
}
|
179 |
+
},
|
180 |
+
"optimizer": {
|
181 |
+
"opt": "adamW",
|
182 |
+
"lr": 1e-05,
|
183 |
+
"opt_betas": [
|
184 |
+
0.9,
|
185 |
+
0.98
|
186 |
+
],
|
187 |
+
"weight_decay": 0.01,
|
188 |
+
"max_grad_norm": 0.7,
|
189 |
+
"different_lr": {
|
190 |
+
"enable": false,
|
191 |
+
"module_names": [],
|
192 |
+
"lr": 1e-05
|
193 |
+
}
|
194 |
+
},
|
195 |
+
"scheduler": {
|
196 |
+
"sched": "cosine",
|
197 |
+
"epochs": 1,
|
198 |
+
"min_lr_multi": 0.01,
|
199 |
+
"warmup_epochs": 0.05
|
200 |
+
},
|
201 |
+
"evaluate": false,
|
202 |
+
"deep_fusion": false,
|
203 |
+
"evaluation": {
|
204 |
+
"eval_frame_ensemble": "concat",
|
205 |
+
"eval_x_only": false,
|
206 |
+
"k_test": 128,
|
207 |
+
"eval_offload": true
|
208 |
+
},
|
209 |
+
"use_half_precision": true,
|
210 |
+
"use_bf16": true,
|
211 |
+
"gradient_checkpointing": true,
|
212 |
+
"wandb": {
|
213 |
+
"enable": true,
|
214 |
+
"entity": "qingy2019-conker-mobile-inc-",
|
215 |
+
"project": "window_iv2"
|
216 |
+
},
|
217 |
+
"dist_url": "env://",
|
218 |
+
"device": "cuda",
|
219 |
+
"mode": "pt",
|
220 |
+
"output_dir": "scripts/pretraining/clip/B14/B14",
|
221 |
+
"resume": true,
|
222 |
+
"debug": false,
|
223 |
+
"log_freq": 1,
|
224 |
+
"seed": 42,
|
225 |
+
"save_latest": false,
|
226 |
+
"save_iter": 5000,
|
227 |
+
"eval_freq_steps": 1000,
|
228 |
+
"eval_video_repo_id": "qingy2024/backflip_train",
|
229 |
+
"eval_video_filename": "1.mp4",
|
230 |
+
"eval_plot_output_dir": "scripts/pretraining/clip/B14/cosine_sim_graphs",
|
231 |
+
"auto_resume": true,
|
232 |
+
"pretrained_path": "",
|
233 |
+
"deepspeed": {
|
234 |
+
"enable": true,
|
235 |
+
"stage": 1
|
236 |
+
},
|
237 |
+
"rank": 0,
|
238 |
+
"world_size": 1,
|
239 |
+
"gpu": 0,
|
240 |
+
"distributed": true,
|
241 |
+
"dist_backend": "nccl",
|
242 |
+
"deepspeed_config": "scripts/pretraining/clip/B14/B14/deepspeed_config.json"
|
243 |
+
}
|
V4.3-ckpt/cosine_sim_graphs/graph_step_0000000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0001000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0002000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0003000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0004000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0005000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0006000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0007000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0008000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0009000.png
ADDED
![]() |
V4.3-ckpt/cosine_sim_graphs/graph_step_0010000.png
ADDED
![]() |