Hugging Face's logo

chancharikm
/
CHAI_SFT_model_8b

Video-Text-to-Text
Transformers
Safetensors
qwen3_vl
image-text-to-text
llama-factory
full
Generated from Trainer
video-language-model
video-captioning
Model card Files
xet
 Community
chancharikm commited on
Commit
6f164cc
·
verified ·
1 Parent(s): 26d7719

Training in progress, step 700, checkpoint

Browse files
Files changed (23) hide show
  1. last-checkpoint/global_step700/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
  2. last-checkpoint/global_step700/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
  3. last-checkpoint/global_step700/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
  4. last-checkpoint/global_step700/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
  5. last-checkpoint/global_step700/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt +3 -0
  6. last-checkpoint/global_step700/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt +3 -0
  7. last-checkpoint/global_step700/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt +3 -0
  8. last-checkpoint/global_step700/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt +3 -0
  9. last-checkpoint/global_step700/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
  10. last-checkpoint/global_step700/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
  11. last-checkpoint/global_step700/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
  12. last-checkpoint/global_step700/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
  13. last-checkpoint/global_step700/zero_pp_rank_4_mp_rank_00_model_states.pt +3 -0
  14. last-checkpoint/global_step700/zero_pp_rank_5_mp_rank_00_model_states.pt +3 -0
  15. last-checkpoint/global_step700/zero_pp_rank_6_mp_rank_00_model_states.pt +3 -0
  16. last-checkpoint/global_step700/zero_pp_rank_7_mp_rank_00_model_states.pt +3 -0
  17. last-checkpoint/latest +1 -1
  18. last-checkpoint/model-00001-of-00004.safetensors +1 -1
  19. last-checkpoint/model-00002-of-00004.safetensors +1 -1
  20. last-checkpoint/model-00003-of-00004.safetensors +1 -1
  21. last-checkpoint/model-00004-of-00004.safetensors +1 -1
  22. last-checkpoint/scheduler.pt +1 -1
  23. last-checkpoint/trainer_state.json +353 -3
last-checkpoint/global_step700/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:f671189f043ff1359c6b1522944efdf619a75bd011b8cf9826473294fe6e49aa
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c826a4ea99a48baddbc314f16ba6b70d80111819aab8145b26ec4260d1980fa
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc6bfcdc42ef72b382081f00b50c227d8cabdca9da51c1219d36787e188b069b
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00d9c95d60c06442da594c838a052a7bc698d7a38e661ab52cdfabe3fb62d1ec
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_4_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3144ac0c1babd6fbbe170f3e6d128aacc4491e99b9f0f1a46b5b9e7f1042ec8a
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_5_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf61ba4ab10f1c2685fadc20811d95641609c43ce3b1a28cd5137816056e8280
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_6_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7dc59e4f594616e8bef878142b6fed63783b5c160810632ec96d9ba408f4bf77
3
+ size 1558836997
last-checkpoint/global_step700/bf16_zero_pp_rank_7_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25b10de9dc9a124bc3a55e9f1ef7e5434d1864532564341c6a8fd98bc9b3e035
3
+ size 1558836997
last-checkpoint/global_step700/zero_pp_rank_0_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f7d2eaa495a913f2ac0fcd672b9e2fedc63b98d0f3f97a2d8de29bf4b4448792
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_1_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d8622b17db5b6b5090bdf619dc710bc677d4700741b7a2aa5b481eefb0f6ac0
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_2_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a2f523cfb96bb252882d9a93a1236cc688c608c91e6461cc7d39122be9f362e
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_3_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:febdc81882b2fe5a999a7b4b2c72bc5c9eb727548767173cc21b3a95a41c008c
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_4_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e276d0a64d79ee69a68864870dff9a102dd9e4376a1ef01b92bb819e76d299d3
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_5_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:198803e0763e11b2f8160d537de3208ff9c60705918517559aa735b16406f0ca
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_6_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f03bca29f89637e2505c9d8d358a7ef4ffdd25a8fea9cd296667e2f4d2206e18
3
+ size 14663005
last-checkpoint/global_step700/zero_pp_rank_7_mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf140a92bb60bda69a13d477f905ac1a1e43cf1d211fc3115e9fdfc7b1773a71
3
+ size 14663005
last-checkpoint/latest CHANGED
@@ -1 +1 @@
1
- global_step650
 
1
+ global_step700
last-checkpoint/model-00001-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0bb3a8341b1a9b9674c51c2e5bd127f72364fc6077156a65fa92c6d6c04a26c2
3
  size 4998056552
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4801dfb1de5603d9fd86e66233f71fd69b7514a405be533d0560a10ac2232701
3
  size 4998056552
last-checkpoint/model-00002-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fed83fd46c620951989793841cded75d428819c2c750477457386f990b1757b2
3
  size 4915962464
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ae2e98b4e0719324774a26fb6680623e8fc7671467315c0748ef1e7d0b4119ba
3
  size 4915962464
last-checkpoint/model-00003-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bc20ce32671487f996e72de9a0ecbeeb481a41f58e2030c6aa9b866ad421209a
3
  size 4915962496
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65839ff8d08fd3e1453b2209252309b7ea1b5d1d54c3d0005a4d9ff668c67d99
3
  size 4915962496
last-checkpoint/model-00004-of-00004.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f366cca33f2493cd496b77dbcbb9450395ea20e95d81127cfd977bb7d5c84296
3
  size 2704357976
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e606573eedf3be0faded56cd50dd9d1206aada29ef5973f67584c3bbf371fc0d
3
  size 2704357976
last-checkpoint/scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d506b2ee8908d92a187e5fac6c90282f80f4f9ffeb0d860e5f3e7a1fe7b59640
3
  size 1465
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52c60310c79aedc98d1b85ec51f9670b286af151f72286d16d7d60189e2aebd3
3
  size 1465
last-checkpoint/trainer_state.json CHANGED
@@ -2,9 +2,9 @@
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
- "epoch": 2.632860040567951,
6
  "eval_steps": 500,
7
- "global_step": 650,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
@@ -4558,6 +4558,356 @@
4558
  "learning_rate": 1.2499710658584296e-06,
4559
  "loss": 0.0108,
4560
  "step": 650
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4561
  }
4562
  ],
4563
  "logging_steps": 1,
@@ -4577,7 +4927,7 @@
4577
  "attributes": {}
4578
  }
4579
  },
4580
- "total_flos": 4402837420769280.0,
4581
  "train_batch_size": 10,
4582
  "trial_name": null,
4583
  "trial_params": null
 
2
  "best_global_step": null,
3
  "best_metric": null,
4
  "best_model_checkpoint": null,
5
+ "epoch": 2.8356997971602436,
6
  "eval_steps": 500,
7
+ "global_step": 700,
8
  "is_hyper_param_search": false,
9
  "is_local_process_zero": true,
10
  "is_world_process_zero": true,
 
4558
  "learning_rate": 1.2499710658584296e-06,
4559
  "loss": 0.0108,
4560
  "step": 650
4561
+ },
4562
+ {
4563
+ "epoch": 2.636916835699797,
4564
+ "grad_norm": 0.31247533393817645,
4565
+ "learning_rate": 1.2233190101337644e-06,
4566
+ "loss": 0.0126,
4567
+ "step": 651
4568
+ },
4569
+ {
4570
+ "epoch": 2.640973630831643,
4571
+ "grad_norm": 0.23703695879461487,
4572
+ "learning_rate": 1.1969420812583603e-06,
4573
+ "loss": 0.0113,
4574
+ "step": 652
4575
+ },
4576
+ {
4577
+ "epoch": 2.6450304259634887,
4578
+ "grad_norm": 0.29797798103465356,
4579
+ "learning_rate": 1.1708408059919901e-06,
4580
+ "loss": 0.0139,
4581
+ "step": 653
4582
+ },
4583
+ {
4584
+ "epoch": 2.6490872210953347,
4585
+ "grad_norm": 3.139650970793164,
4586
+ "learning_rate": 1.145015705589491e-06,
4587
+ "loss": 0.0112,
4588
+ "step": 654
4589
+ },
4590
+ {
4591
+ "epoch": 2.6531440162271807,
4592
+ "grad_norm": 0.2405246638894348,
4593
+ "learning_rate": 1.119467295790363e-06,
4594
+ "loss": 0.011,
4595
+ "step": 655
4596
+ },
4597
+ {
4598
+ "epoch": 2.6572008113590266,
4599
+ "grad_norm": 0.24093581404167375,
4600
+ "learning_rate": 1.0941960868084549e-06,
4601
+ "loss": 0.0108,
4602
+ "step": 656
4603
+ },
4604
+ {
4605
+ "epoch": 2.661257606490872,
4606
+ "grad_norm": 1.4166321141255014,
4607
+ "learning_rate": 1.06920258332179e-06,
4608
+ "loss": 0.0092,
4609
+ "step": 657
4610
+ },
4611
+ {
4612
+ "epoch": 2.665314401622718,
4613
+ "grad_norm": 0.2516080027548331,
4614
+ "learning_rate": 1.0444872844624804e-06,
4615
+ "loss": 0.0091,
4616
+ "step": 658
4617
+ },
4618
+ {
4619
+ "epoch": 2.6693711967545637,
4620
+ "grad_norm": 0.27372507155820414,
4621
+ "learning_rate": 1.0200506838067558e-06,
4622
+ "loss": 0.0161,
4623
+ "step": 659
4624
+ },
4625
+ {
4626
+ "epoch": 2.6734279918864097,
4627
+ "grad_norm": 0.22462772010565904,
4628
+ "learning_rate": 9.958932693651168e-07,
4629
+ "loss": 0.0085,
4630
+ "step": 660
4631
+ },
4632
+ {
4633
+ "epoch": 2.6774847870182557,
4634
+ "grad_norm": 0.19482170746221347,
4635
+ "learning_rate": 9.720155235725753e-07,
4636
+ "loss": 0.0077,
4637
+ "step": 661
4638
+ },
4639
+ {
4640
+ "epoch": 2.6815415821501016,
4641
+ "grad_norm": 0.21706494083610625,
4642
+ "learning_rate": 9.484179232790391e-07,
4643
+ "loss": 0.0071,
4644
+ "step": 662
4645
+ },
4646
+ {
4647
+ "epoch": 2.685598377281947,
4648
+ "grad_norm": 0.2700405708098266,
4649
+ "learning_rate": 9.251009397397663e-07,
4650
+ "loss": 0.0096,
4651
+ "step": 663
4652
+ },
4653
+ {
4654
+ "epoch": 2.689655172413793,
4655
+ "grad_norm": 0.20416894288766446,
4656
+ "learning_rate": 9.020650386059731e-07,
4657
+ "loss": 0.0084,
4658
+ "step": 664
4659
+ },
4660
+ {
4661
+ "epoch": 2.6937119675456387,
4662
+ "grad_norm": 0.2977765225079465,
4663
+ "learning_rate": 8.793106799155187e-07,
4664
+ "loss": 0.0122,
4665
+ "step": 665
4666
+ },
4667
+ {
4668
+ "epoch": 2.6977687626774847,
4669
+ "grad_norm": 0.39111551936936895,
4670
+ "learning_rate": 8.568383180837369e-07,
4671
+ "loss": 0.0061,
4672
+ "step": 666
4673
+ },
4674
+ {
4675
+ "epoch": 2.7018255578093306,
4676
+ "grad_norm": 0.23369929762822697,
4677
+ "learning_rate": 8.346484018943429e-07,
4678
+ "loss": 0.0071,
4679
+ "step": 667
4680
+ },
4681
+ {
4682
+ "epoch": 2.7058823529411766,
4683
+ "grad_norm": 0.2529908334984344,
4684
+ "learning_rate": 8.127413744904805e-07,
4685
+ "loss": 0.0109,
4686
+ "step": 668
4687
+ },
4688
+ {
4689
+ "epoch": 2.709939148073022,
4690
+ "grad_norm": 0.19978914969994244,
4691
+ "learning_rate": 7.911176733658709e-07,
4692
+ "loss": 0.0071,
4693
+ "step": 669
4694
+ },
4695
+ {
4696
+ "epoch": 2.713995943204868,
4697
+ "grad_norm": 0.35516946776094027,
4698
+ "learning_rate": 7.697777303560699e-07,
4699
+ "loss": 0.0076,
4700
+ "step": 670
4701
+ },
4702
+ {
4703
+ "epoch": 2.718052738336714,
4704
+ "grad_norm": 0.32064994090406973,
4705
+ "learning_rate": 7.487219716298643e-07,
4706
+ "loss": 0.0078,
4707
+ "step": 671
4708
+ },
4709
+ {
4710
+ "epoch": 2.7221095334685597,
4711
+ "grad_norm": 0.17610583979954028,
4712
+ "learning_rate": 7.279508176807259e-07,
4713
+ "loss": 0.007,
4714
+ "step": 672
4715
+ },
4716
+ {
4717
+ "epoch": 2.7261663286004056,
4718
+ "grad_norm": 0.1879019239049559,
4719
+ "learning_rate": 7.074646833184495e-07,
4720
+ "loss": 0.0077,
4721
+ "step": 673
4722
+ },
4723
+ {
4724
+ "epoch": 2.7302231237322516,
4725
+ "grad_norm": 0.2508655898096489,
4726
+ "learning_rate": 6.872639776608464e-07,
4727
+ "loss": 0.012,
4728
+ "step": 674
4729
+ },
4730
+ {
4731
+ "epoch": 2.7342799188640976,
4732
+ "grad_norm": 0.20939981063074065,
4733
+ "learning_rate": 6.673491041255874e-07,
4734
+ "loss": 0.0072,
4735
+ "step": 675
4736
+ },
4737
+ {
4738
+ "epoch": 2.738336713995943,
4739
+ "grad_norm": 0.2524661562844755,
4740
+ "learning_rate": 6.477204604221298e-07,
4741
+ "loss": 0.0089,
4742
+ "step": 676
4743
+ },
4744
+ {
4745
+ "epoch": 2.742393509127789,
4746
+ "grad_norm": 0.18123227676882245,
4747
+ "learning_rate": 6.283784385437963e-07,
4748
+ "loss": 0.0074,
4749
+ "step": 677
4750
+ },
4751
+ {
4752
+ "epoch": 2.7464503042596347,
4753
+ "grad_norm": 0.28169632716010695,
4754
+ "learning_rate": 6.093234247599317e-07,
4755
+ "loss": 0.0115,
4756
+ "step": 678
4757
+ },
4758
+ {
4759
+ "epoch": 2.7505070993914806,
4760
+ "grad_norm": 0.24112096399796779,
4761
+ "learning_rate": 5.905557996081906e-07,
4762
+ "loss": 0.0104,
4763
+ "step": 679
4764
+ },
4765
+ {
4766
+ "epoch": 2.7545638945233266,
4767
+ "grad_norm": 1.891213001719549,
4768
+ "learning_rate": 5.720759378869455e-07,
4769
+ "loss": 0.0108,
4770
+ "step": 680
4771
+ },
4772
+ {
4773
+ "epoch": 2.7586206896551726,
4774
+ "grad_norm": 0.2143920645162132,
4775
+ "learning_rate": 5.538842086477891e-07,
4776
+ "loss": 0.0073,
4777
+ "step": 681
4778
+ },
4779
+ {
4780
+ "epoch": 2.762677484787018,
4781
+ "grad_norm": 0.3298995677537989,
4782
+ "learning_rate": 5.359809751881834e-07,
4783
+ "loss": 0.0114,
4784
+ "step": 682
4785
+ },
4786
+ {
4787
+ "epoch": 2.766734279918864,
4788
+ "grad_norm": 0.18127821937547067,
4789
+ "learning_rate": 5.18366595044184e-07,
4790
+ "loss": 0.0059,
4791
+ "step": 683
4792
+ },
4793
+ {
4794
+ "epoch": 2.77079107505071,
4795
+ "grad_norm": 0.3580659253774967,
4796
+ "learning_rate": 5.010414199833097e-07,
4797
+ "loss": 0.0134,
4798
+ "step": 684
4799
+ },
4800
+ {
4801
+ "epoch": 2.7748478701825556,
4802
+ "grad_norm": 0.20714170407676127,
4803
+ "learning_rate": 4.840057959975169e-07,
4804
+ "loss": 0.0072,
4805
+ "step": 685
4806
+ },
4807
+ {
4808
+ "epoch": 2.7789046653144016,
4809
+ "grad_norm": 0.2277337899414502,
4810
+ "learning_rate": 4.6726006329629645e-07,
4811
+ "loss": 0.0101,
4812
+ "step": 686
4813
+ },
4814
+ {
4815
+ "epoch": 2.7829614604462476,
4816
+ "grad_norm": 0.19613471713041244,
4817
+ "learning_rate": 4.5080455629986475e-07,
4818
+ "loss": 0.0047,
4819
+ "step": 687
4820
+ },
4821
+ {
4822
+ "epoch": 2.7870182555780936,
4823
+ "grad_norm": 0.2823039689966462,
4824
+ "learning_rate": 4.3463960363249286e-07,
4825
+ "loss": 0.0115,
4826
+ "step": 688
4827
+ },
4828
+ {
4829
+ "epoch": 2.791075050709939,
4830
+ "grad_norm": 0.20345163681925613,
4831
+ "learning_rate": 4.187655281159536e-07,
4832
+ "loss": 0.0074,
4833
+ "step": 689
4834
+ },
4835
+ {
4836
+ "epoch": 2.795131845841785,
4837
+ "grad_norm": 0.29840307095638663,
4838
+ "learning_rate": 4.031826467630545e-07,
4839
+ "loss": 0.01,
4840
+ "step": 690
4841
+ },
4842
+ {
4843
+ "epoch": 2.7991886409736306,
4844
+ "grad_norm": 0.1860689908066093,
4845
+ "learning_rate": 3.8789127077132336e-07,
4846
+ "loss": 0.0055,
4847
+ "step": 691
4848
+ },
4849
+ {
4850
+ "epoch": 2.8032454361054766,
4851
+ "grad_norm": 0.3478181689879201,
4852
+ "learning_rate": 3.7289170551678633e-07,
4853
+ "loss": 0.0118,
4854
+ "step": 692
4855
+ },
4856
+ {
4857
+ "epoch": 2.8073022312373226,
4858
+ "grad_norm": 0.22939108780939063,
4859
+ "learning_rate": 3.581842505478744e-07,
4860
+ "loss": 0.0112,
4861
+ "step": 693
4862
+ },
4863
+ {
4864
+ "epoch": 2.8113590263691686,
4865
+ "grad_norm": 0.3032587745992294,
4866
+ "learning_rate": 3.437691995794301e-07,
4867
+ "loss": 0.0124,
4868
+ "step": 694
4869
+ },
4870
+ {
4871
+ "epoch": 2.815415821501014,
4872
+ "grad_norm": 0.2094386636776182,
4873
+ "learning_rate": 3.2964684048685523e-07,
4874
+ "loss": 0.0073,
4875
+ "step": 695
4876
+ },
4877
+ {
4878
+ "epoch": 2.81947261663286,
4879
+ "grad_norm": 0.24925291522480858,
4880
+ "learning_rate": 3.15817455300354e-07,
4881
+ "loss": 0.01,
4882
+ "step": 696
4883
+ },
4884
+ {
4885
+ "epoch": 2.8235294117647056,
4886
+ "grad_norm": 0.22927354124709365,
4887
+ "learning_rate": 3.022813201993041e-07,
4888
+ "loss": 0.0103,
4889
+ "step": 697
4890
+ },
4891
+ {
4892
+ "epoch": 2.8275862068965516,
4893
+ "grad_norm": 0.24523657497451284,
4894
+ "learning_rate": 2.890387055067412e-07,
4895
+ "loss": 0.0092,
4896
+ "step": 698
4897
+ },
4898
+ {
4899
+ "epoch": 2.8316430020283976,
4900
+ "grad_norm": 0.2375204245523175,
4901
+ "learning_rate": 2.7608987568395306e-07,
4902
+ "loss": 0.0116,
4903
+ "step": 699
4904
+ },
4905
+ {
4906
+ "epoch": 2.8356997971602436,
4907
+ "grad_norm": 0.2066202436946509,
4908
+ "learning_rate": 2.6343508932521243e-07,
4909
+ "loss": 0.0086,
4910
+ "step": 700
4911
  }
4912
  ],
4913
  "logging_steps": 1,
 
4927
  "attributes": {}
4928
  }
4929
  },
4930
+ "total_flos": 4740780390875136.0,
4931
  "train_batch_size": 10,
4932
  "trial_name": null,
4933
  "trial_params": null