Upload VNEncoder checkpoints (part 2)
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/trainer_state.json +820 -0
- uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/training_args.bin +3 -0
- uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/vocab.txt +0 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/model.safetensors +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/added_tokens.json +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/bpe.codes +0 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/model.safetensors +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/optimizer.pt +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/rng_state.pth +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/scheduler.pt +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/special_tokens_map.json +9 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/tokenizer_config.json +55 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/trainer_state.json +820 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/training_args.bin +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/vocab.txt +0 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/added_tokens.json +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/bpe.codes +0 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/model.safetensors +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/optimizer.pt +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/rng_state.pth +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/scheduler.pt +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/special_tokens_map.json +9 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/tokenizer_config.json +55 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/trainer_state.json +1154 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/training_args.bin +3 -0
- uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/vocab.txt +0 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/model.safetensors +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/added_tokens.json +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/bpe.codes +0 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/model.safetensors +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/optimizer.pt +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/rng_state.pth +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/scheduler.pt +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/special_tokens_map.json +9 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/tokenizer_config.json +55 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/trainer_state.json +139 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/training_args.bin +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/vocab.txt +0 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/added_tokens.json +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/bpe.codes +0 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/model.safetensors +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/optimizer.pt +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/rng_state.pth +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/scheduler.pt +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/special_tokens_map.json +9 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/tokenizer_config.json +55 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/trainer_state.json +235 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/training_args.bin +3 -0
- viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/vocab.txt +0 -0
- viclickbait2025-phobert-s43-20260811T115543Z-6785dbb2/model.safetensors +3 -0
uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/trainer_state.json
ADDED
|
@@ -0,0 +1,820 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1432,
|
| 3 |
+
"best_metric": 0.8863897339195895,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-1432",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2506,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 13.923125267028809,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.1641,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 14.098784446716309,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 3.9552,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 8.778177261352539,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.6133,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 8.585790634155273,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.1693,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 8.640725135803223,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 2.7232,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 15.87416934967041,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 2.1431,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 18.425439834594727,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 1.7422,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 19.100566864013672,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 1.4329,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 13.058442115783691,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.1717,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 16.418716430664062,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.1043,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 7.262048244476318,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.2553,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 11.152338981628418,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.0303,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 42.430885314941406,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.0677,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 31.819955825805664,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 1.0514,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.9286165508528111,
|
| 113 |
+
"eval_loss": 0.22945000231266022,
|
| 114 |
+
"eval_macro_f1": 0.7417401925195808,
|
| 115 |
+
"eval_runtime": 1.2185,
|
| 116 |
+
"eval_samples_per_second": 1299.108,
|
| 117 |
+
"eval_steps_per_second": 81.246,
|
| 118 |
+
"eval_weighted_f1": 0.919070405172157,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 26.900482177734375,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 0.7684,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 38.06690979003906,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 0.9548,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 12.20052433013916,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 0.7953,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 11.588409423828125,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 1.0354,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 5.870014667510986,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 0.7644,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 105.68302917480469,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.8014,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 32.38567352294922,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 0.7367,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 5.859184741973877,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.786,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 55.12186050415039,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 0.8298,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 25.740800857543945,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.7614,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 1.2646484375,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.9679,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 65.3179702758789,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.7893,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 65.5310287475586,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.7466,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 25.88797950744629,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.8709,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9444093493367025,
|
| 222 |
+
"eval_loss": 0.18904562294483185,
|
| 223 |
+
"eval_macro_f1": 0.8471103759014958,
|
| 224 |
+
"eval_runtime": 1.191,
|
| 225 |
+
"eval_samples_per_second": 1329.172,
|
| 226 |
+
"eval_steps_per_second": 83.126,
|
| 227 |
+
"eval_weighted_f1": 0.9453995843643106,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 15.861852645874023,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.7788,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 14.42852783203125,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.6554,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 12.70154857635498,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.8068,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 0.8662025928497314,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.643,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 36.6900520324707,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.6289,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 5.5895209312438965,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.602,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 0.5466787815093994,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.7399,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 47.00026321411133,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.7353,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 45.91748809814453,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.7371,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 37.02352523803711,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.815,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 3.8665976524353027,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.6528,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 91.49957275390625,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.8526,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 0.9258260130882263,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.7223,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 8.029545783996582,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.7686,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9538850284270373,
|
| 331 |
+
"eval_loss": 0.20316267013549805,
|
| 332 |
+
"eval_macro_f1": 0.8692839892986868,
|
| 333 |
+
"eval_runtime": 1.1814,
|
| 334 |
+
"eval_samples_per_second": 1339.951,
|
| 335 |
+
"eval_steps_per_second": 83.8,
|
| 336 |
+
"eval_weighted_f1": 0.9521650356088963,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 1.0149946212768555,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.5123,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 2.508129596710205,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.5017,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 3.8049352169036865,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.5834,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 5.531970977783203,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.663,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 0.7799978256225586,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.5478,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 5.73989725112915,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.5364,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 42.89474868774414,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.5246,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 10.089348793029785,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.5556,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 1.2348963022232056,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.5962,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 61.587135314941406,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.5316,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 3.362563371658325,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.5797,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 0.3001021146774292,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.4235,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 16.912429809570312,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.5791,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 30.365890502929688,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.4653,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 25.31011962890625,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.5386,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9564118761844599,
|
| 447 |
+
"eval_loss": 0.19586284458637238,
|
| 448 |
+
"eval_macro_f1": 0.8863897339195895,
|
| 449 |
+
"eval_runtime": 1.1682,
|
| 450 |
+
"eval_samples_per_second": 1355.11,
|
| 451 |
+
"eval_steps_per_second": 84.748,
|
| 452 |
+
"eval_weighted_f1": 0.9555825690416305,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 1.1835957765579224,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.4636,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 41.54434585571289,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.5932,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 0.3222101032733917,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.4389,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 11.52665901184082,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6008,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 5.204033851623535,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.4368,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 30.632808685302734,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.4196,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 17.37767219543457,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.3684,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 23.26974105834961,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.6812,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 2.531872272491455,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.7023,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 0.29202693700790405,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.3393,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 6.701529502868652,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.4041,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 54.24859619140625,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.3871,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 0.759128987789154,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.3765,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 0.28017839789390564,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.4092,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9494630448515476,
|
| 556 |
+
"eval_loss": 0.2569068372249603,
|
| 557 |
+
"eval_macro_f1": 0.8430213549125547,
|
| 558 |
+
"eval_runtime": 1.1923,
|
| 559 |
+
"eval_samples_per_second": 1327.642,
|
| 560 |
+
"eval_steps_per_second": 83.03,
|
| 561 |
+
"eval_weighted_f1": 0.9460679533759071,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 14.409866333007812,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.4847,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 39.042423248291016,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.2839,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 3.189486503601074,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.3604,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 0.9908621907234192,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.6025,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 29.93218421936035,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.2045,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 0.18825837969779968,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.1893,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 31.050430297851562,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.4356,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 0.18963059782981873,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.4483,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 3.1391115188598633,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.4364,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 0.20262311398983002,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.3424,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 0.2473168969154358,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.3252,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 99.61351013183594,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.3562,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 69.8361587524414,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.2312,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 1.6530143022537231,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.4532,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9444093493367025,
|
| 665 |
+
"eval_loss": 0.264648973941803,
|
| 666 |
+
"eval_macro_f1": 0.8509371552388153,
|
| 667 |
+
"eval_runtime": 1.1875,
|
| 668 |
+
"eval_samples_per_second": 1333.055,
|
| 669 |
+
"eval_steps_per_second": 83.369,
|
| 670 |
+
"eval_weighted_f1": 0.9443039887730623,
|
| 671 |
+
"step": 2148
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 6.005598320503849,
|
| 675 |
+
"grad_norm": 45.83721160888672,
|
| 676 |
+
"learning_rate": 1.5552451893234017e-05,
|
| 677 |
+
"loss": 0.3795,
|
| 678 |
+
"step": 2150
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 6.07557732680196,
|
| 682 |
+
"grad_norm": 6.040622711181641,
|
| 683 |
+
"learning_rate": 1.5474860335195532e-05,
|
| 684 |
+
"loss": 0.2085,
|
| 685 |
+
"step": 2175
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 6.14555633310007,
|
| 689 |
+
"grad_norm": 1.5008230209350586,
|
| 690 |
+
"learning_rate": 1.5397268777157047e-05,
|
| 691 |
+
"loss": 0.3399,
|
| 692 |
+
"step": 2200
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.215535339398181,
|
| 696 |
+
"grad_norm": 5.875795364379883,
|
| 697 |
+
"learning_rate": 1.531967721911856e-05,
|
| 698 |
+
"loss": 0.3668,
|
| 699 |
+
"step": 2225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 6.285514345696291,
|
| 703 |
+
"grad_norm": 0.2164689004421234,
|
| 704 |
+
"learning_rate": 1.5242085661080076e-05,
|
| 705 |
+
"loss": 0.2451,
|
| 706 |
+
"step": 2250
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.355493351994402,
|
| 710 |
+
"grad_norm": 35.070892333984375,
|
| 711 |
+
"learning_rate": 1.5164494103041589e-05,
|
| 712 |
+
"loss": 0.4848,
|
| 713 |
+
"step": 2275
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.425472358292512,
|
| 717 |
+
"grad_norm": 13.294031143188477,
|
| 718 |
+
"learning_rate": 1.5086902545003106e-05,
|
| 719 |
+
"loss": 0.3581,
|
| 720 |
+
"step": 2300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.495451364590623,
|
| 724 |
+
"grad_norm": 5.004824638366699,
|
| 725 |
+
"learning_rate": 1.5009310986964621e-05,
|
| 726 |
+
"loss": 0.3002,
|
| 727 |
+
"step": 2325
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 6.565430370888733,
|
| 731 |
+
"grad_norm": 7.410371780395508,
|
| 732 |
+
"learning_rate": 1.4931719428926134e-05,
|
| 733 |
+
"loss": 0.3237,
|
| 734 |
+
"step": 2350
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.635409377186844,
|
| 738 |
+
"grad_norm": 49.6166877746582,
|
| 739 |
+
"learning_rate": 1.485412787088765e-05,
|
| 740 |
+
"loss": 0.3872,
|
| 741 |
+
"step": 2375
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.705388383484954,
|
| 745 |
+
"grad_norm": 43.88654708862305,
|
| 746 |
+
"learning_rate": 1.4776536312849163e-05,
|
| 747 |
+
"loss": 0.1795,
|
| 748 |
+
"step": 2400
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 6.7753673897830655,
|
| 752 |
+
"grad_norm": 4.089932918548584,
|
| 753 |
+
"learning_rate": 1.4698944754810678e-05,
|
| 754 |
+
"loss": 0.4171,
|
| 755 |
+
"step": 2425
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 6.845346396081176,
|
| 759 |
+
"grad_norm": 35.39617156982422,
|
| 760 |
+
"learning_rate": 1.4621353196772193e-05,
|
| 761 |
+
"loss": 0.3371,
|
| 762 |
+
"step": 2450
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.9153254023792865,
|
| 766 |
+
"grad_norm": 0.15443751215934753,
|
| 767 |
+
"learning_rate": 1.4543761638733706e-05,
|
| 768 |
+
"loss": 0.2292,
|
| 769 |
+
"step": 2475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 6.985304408677397,
|
| 773 |
+
"grad_norm": 0.48482105135917664,
|
| 774 |
+
"learning_rate": 1.4466170080695221e-05,
|
| 775 |
+
"loss": 0.2385,
|
| 776 |
+
"step": 2500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.0,
|
| 780 |
+
"eval_accuracy": 0.950726468730259,
|
| 781 |
+
"eval_loss": 0.25465235114097595,
|
| 782 |
+
"eval_macro_f1": 0.8692058127987776,
|
| 783 |
+
"eval_runtime": 1.2347,
|
| 784 |
+
"eval_samples_per_second": 1282.13,
|
| 785 |
+
"eval_steps_per_second": 80.184,
|
| 786 |
+
"eval_weighted_f1": 0.9505632312183229,
|
| 787 |
+
"step": 2506
|
| 788 |
+
}
|
| 789 |
+
],
|
| 790 |
+
"logging_steps": 25,
|
| 791 |
+
"max_steps": 7160,
|
| 792 |
+
"num_input_tokens_seen": 0,
|
| 793 |
+
"num_train_epochs": 20,
|
| 794 |
+
"save_steps": 500,
|
| 795 |
+
"stateful_callbacks": {
|
| 796 |
+
"EarlyStoppingCallback": {
|
| 797 |
+
"args": {
|
| 798 |
+
"early_stopping_patience": 3,
|
| 799 |
+
"early_stopping_threshold": 0.0
|
| 800 |
+
},
|
| 801 |
+
"attributes": {
|
| 802 |
+
"early_stopping_patience_counter": 3
|
| 803 |
+
}
|
| 804 |
+
},
|
| 805 |
+
"TrainerControl": {
|
| 806 |
+
"args": {
|
| 807 |
+
"should_epoch_stop": false,
|
| 808 |
+
"should_evaluate": false,
|
| 809 |
+
"should_log": false,
|
| 810 |
+
"should_save": true,
|
| 811 |
+
"should_training_stop": true
|
| 812 |
+
},
|
| 813 |
+
"attributes": {}
|
| 814 |
+
}
|
| 815 |
+
},
|
| 816 |
+
"total_flos": 0.0,
|
| 817 |
+
"train_batch_size": 8,
|
| 818 |
+
"trial_name": null,
|
| 819 |
+
"trial_params": null
|
| 820 |
+
}
|
uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b7fcee69fce83e757357ac2f61a1b4b23e79ce79e281be6566d03aacdd295850
|
| 3 |
+
size 5560
|
uit_vsfc-phobert-s43-20260811T113512Z-d606ec95/trainer/checkpoint-2506/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8125ebf17d562047348258e4788c81f9036534c031af42777c2b79382b4133ee
|
| 3 |
+
size 540026404
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<mask>": 64000
|
| 3 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/bpe.codes
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b1c9df97b58678355bcd0ddeda8169817b48e6580244446c06a75c11564d560
|
| 3 |
+
size 540026436
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffdc5aa681ea67646849c9d505e0d5b6631fc6d9d1f1c35f6dce282d48b49268
|
| 3 |
+
size 1075446330
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4f45ce3ecdb182573b50b4f807a34d3886973cdea6ebf992e061bcd94dfc491
|
| 3 |
+
size 14244
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47bd5208186c890baffd5cbaf1c40acb62204621d2c3491c09c0c6bc53644992
|
| 3 |
+
size 1064
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/special_tokens_map.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": "<mask>",
|
| 6 |
+
"pad_token": "<pad>",
|
| 7 |
+
"sep_token": "</s>",
|
| 8 |
+
"unk_token": "<unk>"
|
| 9 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"64000": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 256,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "PhobertTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/trainer_state.json
ADDED
|
@@ -0,0 +1,820 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2506,
|
| 3 |
+
"best_metric": 0.8662234550886146,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2506,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 15.890106201171875,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.7624,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 15.995828628540039,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.4561,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 14.525657653808594,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.9092,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 12.575777053833008,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4007,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 11.675627708435059,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 2.9089,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 11.703021049499512,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 2.3688,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 16.696861267089844,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 1.8084,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 9.186975479125977,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 1.3878,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 44.557212829589844,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.2588,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 38.275447845458984,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.1963,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 11.042539596557617,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.2562,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 59.32037353515625,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.04,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 30.43923568725586,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.0702,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 4.472817420959473,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 0.8881,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.934301958307012,
|
| 113 |
+
"eval_loss": 0.22739678621292114,
|
| 114 |
+
"eval_macro_f1": 0.730669013725883,
|
| 115 |
+
"eval_runtime": 1.263,
|
| 116 |
+
"eval_samples_per_second": 1253.38,
|
| 117 |
+
"eval_steps_per_second": 78.386,
|
| 118 |
+
"eval_weighted_f1": 0.9227602628237189,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 23.841754913330078,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 0.8311,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 27.91180992126465,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 0.9532,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 14.792951583862305,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 0.9974,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 23.817411422729492,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.7487,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 12.263633728027344,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 0.8817,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 0.9602521657943726,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.7935,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 32.26633071899414,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 0.9758,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 21.639888763427734,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.7991,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 39.89500045776367,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 0.9374,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 45.460105895996094,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.7271,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 40.79924011230469,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.9028,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 10.432962417602539,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9458,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 64.54894256591797,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.8605,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 4.725041389465332,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.7906,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9368288060644346,
|
| 222 |
+
"eval_loss": 0.2617494761943817,
|
| 223 |
+
"eval_macro_f1": 0.8142851984654614,
|
| 224 |
+
"eval_runtime": 1.2987,
|
| 225 |
+
"eval_samples_per_second": 1218.878,
|
| 226 |
+
"eval_steps_per_second": 76.228,
|
| 227 |
+
"eval_weighted_f1": 0.9319904145351771,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 34.45173263549805,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.7407,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 56.12771224975586,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.7946,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 15.157228469848633,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.6901,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 5.805321216583252,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.7918,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 7.1828179359436035,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.7319,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 26.249588012695312,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.6547,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 34.88176345825195,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.6213,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 18.288930892944336,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8678,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 22.59798812866211,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.5624,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 122.50627899169922,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.7674,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 12.144521713256836,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.8799,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 30.38553810119629,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.6182,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 37.2827262878418,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.7315,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 5.852736949920654,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.5628,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9437776373973468,
|
| 331 |
+
"eval_loss": 0.19977933168411255,
|
| 332 |
+
"eval_macro_f1": 0.8531590770230153,
|
| 333 |
+
"eval_runtime": 1.3212,
|
| 334 |
+
"eval_samples_per_second": 1198.164,
|
| 335 |
+
"eval_steps_per_second": 74.933,
|
| 336 |
+
"eval_weighted_f1": 0.9448234011581389,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 37.21408462524414,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.7366,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 6.871343612670898,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.5389,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 8.963614463806152,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.6911,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 20.769990921020508,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.6363,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 33.4630126953125,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.5506,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 4.213133335113525,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.5672,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 7.317019462585449,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.514,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 19.837966918945312,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.6475,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 18.455398559570312,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.6985,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 1.2877476215362549,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.7145,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 46.33899688720703,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.4402,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 0.3435637652873993,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.5126,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 1.4445362091064453,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.5023,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 0.9608930945396423,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.4957,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 10.732783317565918,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.58,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9469361970941251,
|
| 447 |
+
"eval_loss": 0.24154046177864075,
|
| 448 |
+
"eval_macro_f1": 0.8294877169586066,
|
| 449 |
+
"eval_runtime": 1.2972,
|
| 450 |
+
"eval_samples_per_second": 1220.359,
|
| 451 |
+
"eval_steps_per_second": 76.321,
|
| 452 |
+
"eval_weighted_f1": 0.9428227509673922,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 5.823998928070068,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.4628,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 1.7952710390090942,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.5368,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 1.0572470426559448,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.5197,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 8.489173889160156,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6243,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 45.122440338134766,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.3062,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 14.028529167175293,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.3844,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 0.44512128829956055,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.4883,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 20.564664840698242,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.4162,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 2.2896320819854736,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.4814,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 2.053428888320923,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.3705,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 41.13652801513672,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.4588,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 3.011770248413086,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.4399,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 15.254034042358398,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.247,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 14.806401252746582,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.5831,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9406190777005685,
|
| 556 |
+
"eval_loss": 0.26180794835090637,
|
| 557 |
+
"eval_macro_f1": 0.8366219283418204,
|
| 558 |
+
"eval_runtime": 1.2905,
|
| 559 |
+
"eval_samples_per_second": 1226.621,
|
| 560 |
+
"eval_steps_per_second": 76.712,
|
| 561 |
+
"eval_weighted_f1": 0.9416365846055232,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 49.11066436767578,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.4847,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 3.138051748275757,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.3709,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 68.00996398925781,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.4448,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 0.19001732766628265,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.2663,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 0.25441697239875793,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.3225,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 17.98102378845215,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.6131,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 2.8782224655151367,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.4778,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 48.173675537109375,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.3094,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 78.5799789428711,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.4037,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 0.39592620730400085,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.3915,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 0.23372972011566162,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.3545,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 0.20506474375724792,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.3706,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 7.170314311981201,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.2673,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 0.43047407269477844,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.4991,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9538850284270373,
|
| 665 |
+
"eval_loss": 0.24408170580863953,
|
| 666 |
+
"eval_macro_f1": 0.8642730226359241,
|
| 667 |
+
"eval_runtime": 1.2865,
|
| 668 |
+
"eval_samples_per_second": 1230.458,
|
| 669 |
+
"eval_steps_per_second": 76.952,
|
| 670 |
+
"eval_weighted_f1": 0.9517727315836694,
|
| 671 |
+
"step": 2148
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 6.005598320503849,
|
| 675 |
+
"grad_norm": 26.87923812866211,
|
| 676 |
+
"learning_rate": 1.5552451893234017e-05,
|
| 677 |
+
"loss": 0.2136,
|
| 678 |
+
"step": 2150
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 6.07557732680196,
|
| 682 |
+
"grad_norm": 200.5290985107422,
|
| 683 |
+
"learning_rate": 1.5474860335195532e-05,
|
| 684 |
+
"loss": 0.2688,
|
| 685 |
+
"step": 2175
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 6.14555633310007,
|
| 689 |
+
"grad_norm": 0.23392556607723236,
|
| 690 |
+
"learning_rate": 1.5397268777157047e-05,
|
| 691 |
+
"loss": 0.4127,
|
| 692 |
+
"step": 2200
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.215535339398181,
|
| 696 |
+
"grad_norm": 2.0901408195495605,
|
| 697 |
+
"learning_rate": 1.531967721911856e-05,
|
| 698 |
+
"loss": 0.3672,
|
| 699 |
+
"step": 2225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 6.285514345696291,
|
| 703 |
+
"grad_norm": 85.51931762695312,
|
| 704 |
+
"learning_rate": 1.5242085661080076e-05,
|
| 705 |
+
"loss": 0.3108,
|
| 706 |
+
"step": 2250
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.355493351994402,
|
| 710 |
+
"grad_norm": 120.88185119628906,
|
| 711 |
+
"learning_rate": 1.5164494103041589e-05,
|
| 712 |
+
"loss": 0.4208,
|
| 713 |
+
"step": 2275
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.425472358292512,
|
| 717 |
+
"grad_norm": 0.24016176164150238,
|
| 718 |
+
"learning_rate": 1.5086902545003106e-05,
|
| 719 |
+
"loss": 0.2161,
|
| 720 |
+
"step": 2300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.495451364590623,
|
| 724 |
+
"grad_norm": 10.961566925048828,
|
| 725 |
+
"learning_rate": 1.5009310986964621e-05,
|
| 726 |
+
"loss": 0.2972,
|
| 727 |
+
"step": 2325
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 6.565430370888733,
|
| 731 |
+
"grad_norm": 23.026823043823242,
|
| 732 |
+
"learning_rate": 1.4931719428926134e-05,
|
| 733 |
+
"loss": 0.2287,
|
| 734 |
+
"step": 2350
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.635409377186844,
|
| 738 |
+
"grad_norm": 71.0625,
|
| 739 |
+
"learning_rate": 1.485412787088765e-05,
|
| 740 |
+
"loss": 0.3265,
|
| 741 |
+
"step": 2375
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.705388383484954,
|
| 745 |
+
"grad_norm": 119.8893814086914,
|
| 746 |
+
"learning_rate": 1.4776536312849163e-05,
|
| 747 |
+
"loss": 0.396,
|
| 748 |
+
"step": 2400
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 6.7753673897830655,
|
| 752 |
+
"grad_norm": 34.25016784667969,
|
| 753 |
+
"learning_rate": 1.4698944754810678e-05,
|
| 754 |
+
"loss": 0.4438,
|
| 755 |
+
"step": 2425
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 6.845346396081176,
|
| 759 |
+
"grad_norm": 0.3095424771308899,
|
| 760 |
+
"learning_rate": 1.4621353196772193e-05,
|
| 761 |
+
"loss": 0.3035,
|
| 762 |
+
"step": 2450
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.9153254023792865,
|
| 766 |
+
"grad_norm": 1.8962700366973877,
|
| 767 |
+
"learning_rate": 1.4543761638733706e-05,
|
| 768 |
+
"loss": 0.2966,
|
| 769 |
+
"step": 2475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 6.985304408677397,
|
| 773 |
+
"grad_norm": 18.388277053833008,
|
| 774 |
+
"learning_rate": 1.4466170080695221e-05,
|
| 775 |
+
"loss": 0.2788,
|
| 776 |
+
"step": 2500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.0,
|
| 780 |
+
"eval_accuracy": 0.9513581806696146,
|
| 781 |
+
"eval_loss": 0.2554718554019928,
|
| 782 |
+
"eval_macro_f1": 0.8662234550886146,
|
| 783 |
+
"eval_runtime": 1.3015,
|
| 784 |
+
"eval_samples_per_second": 1216.247,
|
| 785 |
+
"eval_steps_per_second": 76.063,
|
| 786 |
+
"eval_weighted_f1": 0.9503012319701283,
|
| 787 |
+
"step": 2506
|
| 788 |
+
}
|
| 789 |
+
],
|
| 790 |
+
"logging_steps": 25,
|
| 791 |
+
"max_steps": 7160,
|
| 792 |
+
"num_input_tokens_seen": 0,
|
| 793 |
+
"num_train_epochs": 20,
|
| 794 |
+
"save_steps": 500,
|
| 795 |
+
"stateful_callbacks": {
|
| 796 |
+
"EarlyStoppingCallback": {
|
| 797 |
+
"args": {
|
| 798 |
+
"early_stopping_patience": 3,
|
| 799 |
+
"early_stopping_threshold": 0.0
|
| 800 |
+
},
|
| 801 |
+
"attributes": {
|
| 802 |
+
"early_stopping_patience_counter": 0
|
| 803 |
+
}
|
| 804 |
+
},
|
| 805 |
+
"TrainerControl": {
|
| 806 |
+
"args": {
|
| 807 |
+
"should_epoch_stop": false,
|
| 808 |
+
"should_evaluate": false,
|
| 809 |
+
"should_log": false,
|
| 810 |
+
"should_save": true,
|
| 811 |
+
"should_training_stop": false
|
| 812 |
+
},
|
| 813 |
+
"attributes": {}
|
| 814 |
+
}
|
| 815 |
+
},
|
| 816 |
+
"total_flos": 0.0,
|
| 817 |
+
"train_batch_size": 8,
|
| 818 |
+
"trial_name": null,
|
| 819 |
+
"trial_params": null
|
| 820 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1d15437a12ab74f1b55f89335b5998bd3326c84823b685185c3f07be37b95cb
|
| 3 |
+
size 5560
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<mask>": 64000
|
| 3 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/bpe.codes
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0f7d217c2cab193713744004ebc712e3e121fe275f6f9fc332f4f82aaa303fdc
|
| 3 |
+
size 540026436
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c0b71d38b4fc323ed59f01e34b9e000e97089a063d21a1b337be15aea8dc84fc
|
| 3 |
+
size 1075446330
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdcaf34bedb0ca1abe42a0379b334907c70f2ee0eb9aeaaf5231009eecbec126
|
| 3 |
+
size 14244
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c35ea2d3d14bcfc37d4e8585ed245d16366ae10751c1f309a7a4b190ce29ea91
|
| 3 |
+
size 1064
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/special_tokens_map.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": "<mask>",
|
| 6 |
+
"pad_token": "<pad>",
|
| 7 |
+
"sep_token": "</s>",
|
| 8 |
+
"unk_token": "<unk>"
|
| 9 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"64000": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 256,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "PhobertTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/trainer_state.json
ADDED
|
@@ -0,0 +1,1154 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2506,
|
| 3 |
+
"best_metric": 0.8662234550886146,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-2506",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3580,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.06997900629811056,
|
| 14 |
+
"grad_norm": 15.890106201171875,
|
| 15 |
+
"learning_rate": 6.703910614525141e-07,
|
| 16 |
+
"loss": 4.7624,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13995801259622112,
|
| 21 |
+
"grad_norm": 15.995828628540039,
|
| 22 |
+
"learning_rate": 1.3687150837988828e-06,
|
| 23 |
+
"loss": 4.4561,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2099370188943317,
|
| 28 |
+
"grad_norm": 14.525657653808594,
|
| 29 |
+
"learning_rate": 2.067039106145252e-06,
|
| 30 |
+
"loss": 3.9092,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.27991602519244224,
|
| 35 |
+
"grad_norm": 12.575777053833008,
|
| 36 |
+
"learning_rate": 2.7653631284916204e-06,
|
| 37 |
+
"loss": 3.4007,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.34989503149055284,
|
| 42 |
+
"grad_norm": 11.675627708435059,
|
| 43 |
+
"learning_rate": 3.4636871508379893e-06,
|
| 44 |
+
"loss": 2.9089,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.4198740377886634,
|
| 49 |
+
"grad_norm": 11.703021049499512,
|
| 50 |
+
"learning_rate": 4.1620111731843575e-06,
|
| 51 |
+
"loss": 2.3688,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.489853044086774,
|
| 56 |
+
"grad_norm": 16.696861267089844,
|
| 57 |
+
"learning_rate": 4.860335195530727e-06,
|
| 58 |
+
"loss": 1.8084,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.5598320503848845,
|
| 63 |
+
"grad_norm": 9.186975479125977,
|
| 64 |
+
"learning_rate": 5.558659217877096e-06,
|
| 65 |
+
"loss": 1.3878,
|
| 66 |
+
"step": 200
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.6298110566829951,
|
| 70 |
+
"grad_norm": 44.557212829589844,
|
| 71 |
+
"learning_rate": 6.256983240223464e-06,
|
| 72 |
+
"loss": 1.2588,
|
| 73 |
+
"step": 225
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6997900629811057,
|
| 77 |
+
"grad_norm": 38.275447845458984,
|
| 78 |
+
"learning_rate": 6.9553072625698325e-06,
|
| 79 |
+
"loss": 1.1963,
|
| 80 |
+
"step": 250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7697690692792163,
|
| 84 |
+
"grad_norm": 11.042539596557617,
|
| 85 |
+
"learning_rate": 7.653631284916202e-06,
|
| 86 |
+
"loss": 1.2562,
|
| 87 |
+
"step": 275
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.8397480755773268,
|
| 91 |
+
"grad_norm": 59.32037353515625,
|
| 92 |
+
"learning_rate": 8.35195530726257e-06,
|
| 93 |
+
"loss": 1.04,
|
| 94 |
+
"step": 300
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.9097270818754374,
|
| 98 |
+
"grad_norm": 30.43923568725586,
|
| 99 |
+
"learning_rate": 9.050279329608939e-06,
|
| 100 |
+
"loss": 1.0702,
|
| 101 |
+
"step": 325
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.979706088173548,
|
| 105 |
+
"grad_norm": 4.472817420959473,
|
| 106 |
+
"learning_rate": 9.748603351955308e-06,
|
| 107 |
+
"loss": 0.8881,
|
| 108 |
+
"step": 350
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.0,
|
| 112 |
+
"eval_accuracy": 0.934301958307012,
|
| 113 |
+
"eval_loss": 0.22739678621292114,
|
| 114 |
+
"eval_macro_f1": 0.730669013725883,
|
| 115 |
+
"eval_runtime": 1.263,
|
| 116 |
+
"eval_samples_per_second": 1253.38,
|
| 117 |
+
"eval_steps_per_second": 78.386,
|
| 118 |
+
"eval_weighted_f1": 0.9227602628237189,
|
| 119 |
+
"step": 358
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.0475857242827151,
|
| 123 |
+
"grad_norm": 23.841754913330078,
|
| 124 |
+
"learning_rate": 1.0446927374301676e-05,
|
| 125 |
+
"loss": 0.8311,
|
| 126 |
+
"step": 375
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.1175647305808258,
|
| 130 |
+
"grad_norm": 27.91180992126465,
|
| 131 |
+
"learning_rate": 1.1145251396648044e-05,
|
| 132 |
+
"loss": 0.9532,
|
| 133 |
+
"step": 400
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.1875437368789363,
|
| 137 |
+
"grad_norm": 14.792951583862305,
|
| 138 |
+
"learning_rate": 1.1843575418994415e-05,
|
| 139 |
+
"loss": 0.9974,
|
| 140 |
+
"step": 425
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.2575227431770468,
|
| 144 |
+
"grad_norm": 23.817411422729492,
|
| 145 |
+
"learning_rate": 1.2541899441340784e-05,
|
| 146 |
+
"loss": 0.7487,
|
| 147 |
+
"step": 450
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.3275017494751575,
|
| 151 |
+
"grad_norm": 12.263633728027344,
|
| 152 |
+
"learning_rate": 1.3240223463687152e-05,
|
| 153 |
+
"loss": 0.8817,
|
| 154 |
+
"step": 475
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.397480755773268,
|
| 158 |
+
"grad_norm": 0.9602521657943726,
|
| 159 |
+
"learning_rate": 1.393854748603352e-05,
|
| 160 |
+
"loss": 0.7935,
|
| 161 |
+
"step": 500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.4674597620713785,
|
| 165 |
+
"grad_norm": 32.26633071899414,
|
| 166 |
+
"learning_rate": 1.463687150837989e-05,
|
| 167 |
+
"loss": 0.9758,
|
| 168 |
+
"step": 525
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.5374387683694892,
|
| 172 |
+
"grad_norm": 21.639888763427734,
|
| 173 |
+
"learning_rate": 1.533519553072626e-05,
|
| 174 |
+
"loss": 0.7991,
|
| 175 |
+
"step": 550
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.6074177746675997,
|
| 179 |
+
"grad_norm": 39.89500045776367,
|
| 180 |
+
"learning_rate": 1.6033519553072626e-05,
|
| 181 |
+
"loss": 0.9374,
|
| 182 |
+
"step": 575
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.6773967809657102,
|
| 186 |
+
"grad_norm": 45.460105895996094,
|
| 187 |
+
"learning_rate": 1.6731843575418994e-05,
|
| 188 |
+
"loss": 0.7271,
|
| 189 |
+
"step": 600
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7473757872638207,
|
| 193 |
+
"grad_norm": 40.79924011230469,
|
| 194 |
+
"learning_rate": 1.7430167597765365e-05,
|
| 195 |
+
"loss": 0.9028,
|
| 196 |
+
"step": 625
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.8173547935619314,
|
| 200 |
+
"grad_norm": 10.432962417602539,
|
| 201 |
+
"learning_rate": 1.8128491620111733e-05,
|
| 202 |
+
"loss": 0.9458,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8873337998600421,
|
| 207 |
+
"grad_norm": 64.54894256591797,
|
| 208 |
+
"learning_rate": 1.8826815642458104e-05,
|
| 209 |
+
"loss": 0.8605,
|
| 210 |
+
"step": 675
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.9573128061581526,
|
| 214 |
+
"grad_norm": 4.725041389465332,
|
| 215 |
+
"learning_rate": 1.952513966480447e-05,
|
| 216 |
+
"loss": 0.7906,
|
| 217 |
+
"step": 700
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.0,
|
| 221 |
+
"eval_accuracy": 0.9368288060644346,
|
| 222 |
+
"eval_loss": 0.2617494761943817,
|
| 223 |
+
"eval_macro_f1": 0.8142851984654614,
|
| 224 |
+
"eval_runtime": 1.2987,
|
| 225 |
+
"eval_samples_per_second": 1218.878,
|
| 226 |
+
"eval_steps_per_second": 76.228,
|
| 227 |
+
"eval_weighted_f1": 0.9319904145351771,
|
| 228 |
+
"step": 716
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.0251924422673198,
|
| 232 |
+
"grad_norm": 34.45173263549805,
|
| 233 |
+
"learning_rate": 1.9975170701427685e-05,
|
| 234 |
+
"loss": 0.7407,
|
| 235 |
+
"step": 725
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 2.0951714485654302,
|
| 239 |
+
"grad_norm": 56.12771224975586,
|
| 240 |
+
"learning_rate": 1.98975791433892e-05,
|
| 241 |
+
"loss": 0.7946,
|
| 242 |
+
"step": 750
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 2.1651504548635407,
|
| 246 |
+
"grad_norm": 15.157228469848633,
|
| 247 |
+
"learning_rate": 1.9819987585350715e-05,
|
| 248 |
+
"loss": 0.6901,
|
| 249 |
+
"step": 775
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.2351294611616517,
|
| 253 |
+
"grad_norm": 5.805321216583252,
|
| 254 |
+
"learning_rate": 1.974239602731223e-05,
|
| 255 |
+
"loss": 0.7918,
|
| 256 |
+
"step": 800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.305108467459762,
|
| 260 |
+
"grad_norm": 7.1828179359436035,
|
| 261 |
+
"learning_rate": 1.9664804469273745e-05,
|
| 262 |
+
"loss": 0.7319,
|
| 263 |
+
"step": 825
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 2.3750874737578727,
|
| 267 |
+
"grad_norm": 26.249588012695312,
|
| 268 |
+
"learning_rate": 1.9587212911235257e-05,
|
| 269 |
+
"loss": 0.6547,
|
| 270 |
+
"step": 850
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.445066480055983,
|
| 274 |
+
"grad_norm": 34.88176345825195,
|
| 275 |
+
"learning_rate": 1.9509621353196775e-05,
|
| 276 |
+
"loss": 0.6213,
|
| 277 |
+
"step": 875
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 2.5150454863540936,
|
| 281 |
+
"grad_norm": 18.288930892944336,
|
| 282 |
+
"learning_rate": 1.943202979515829e-05,
|
| 283 |
+
"loss": 0.8678,
|
| 284 |
+
"step": 900
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.585024492652204,
|
| 288 |
+
"grad_norm": 22.59798812866211,
|
| 289 |
+
"learning_rate": 1.9354438237119802e-05,
|
| 290 |
+
"loss": 0.5624,
|
| 291 |
+
"step": 925
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 2.655003498950315,
|
| 295 |
+
"grad_norm": 122.50627899169922,
|
| 296 |
+
"learning_rate": 1.9276846679081317e-05,
|
| 297 |
+
"loss": 0.7674,
|
| 298 |
+
"step": 950
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.7249825052484256,
|
| 302 |
+
"grad_norm": 12.144521713256836,
|
| 303 |
+
"learning_rate": 1.9199255121042832e-05,
|
| 304 |
+
"loss": 0.8799,
|
| 305 |
+
"step": 975
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.794961511546536,
|
| 309 |
+
"grad_norm": 30.38553810119629,
|
| 310 |
+
"learning_rate": 1.9121663563004347e-05,
|
| 311 |
+
"loss": 0.6182,
|
| 312 |
+
"step": 1000
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.8649405178446465,
|
| 316 |
+
"grad_norm": 37.2827262878418,
|
| 317 |
+
"learning_rate": 1.9044072004965862e-05,
|
| 318 |
+
"loss": 0.7315,
|
| 319 |
+
"step": 1025
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 2.934919524142757,
|
| 323 |
+
"grad_norm": 5.852736949920654,
|
| 324 |
+
"learning_rate": 1.8966480446927374e-05,
|
| 325 |
+
"loss": 0.5628,
|
| 326 |
+
"step": 1050
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 3.0,
|
| 330 |
+
"eval_accuracy": 0.9437776373973468,
|
| 331 |
+
"eval_loss": 0.19977933168411255,
|
| 332 |
+
"eval_macro_f1": 0.8531590770230153,
|
| 333 |
+
"eval_runtime": 1.3212,
|
| 334 |
+
"eval_samples_per_second": 1198.164,
|
| 335 |
+
"eval_steps_per_second": 74.933,
|
| 336 |
+
"eval_weighted_f1": 0.9448234011581389,
|
| 337 |
+
"step": 1074
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 3.0027991602519246,
|
| 341 |
+
"grad_norm": 37.21408462524414,
|
| 342 |
+
"learning_rate": 1.888888888888889e-05,
|
| 343 |
+
"loss": 0.7366,
|
| 344 |
+
"step": 1075
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.072778166550035,
|
| 348 |
+
"grad_norm": 6.871343612670898,
|
| 349 |
+
"learning_rate": 1.8811297330850404e-05,
|
| 350 |
+
"loss": 0.5389,
|
| 351 |
+
"step": 1100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 3.1427571728481456,
|
| 355 |
+
"grad_norm": 8.963614463806152,
|
| 356 |
+
"learning_rate": 1.873370577281192e-05,
|
| 357 |
+
"loss": 0.6911,
|
| 358 |
+
"step": 1125
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 3.212736179146256,
|
| 362 |
+
"grad_norm": 20.769990921020508,
|
| 363 |
+
"learning_rate": 1.8656114214773434e-05,
|
| 364 |
+
"loss": 0.6363,
|
| 365 |
+
"step": 1150
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 3.2827151854443666,
|
| 369 |
+
"grad_norm": 33.4630126953125,
|
| 370 |
+
"learning_rate": 1.857852265673495e-05,
|
| 371 |
+
"loss": 0.5506,
|
| 372 |
+
"step": 1175
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.352694191742477,
|
| 376 |
+
"grad_norm": 4.213133335113525,
|
| 377 |
+
"learning_rate": 1.8500931098696464e-05,
|
| 378 |
+
"loss": 0.5672,
|
| 379 |
+
"step": 1200
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 3.422673198040588,
|
| 383 |
+
"grad_norm": 7.317019462585449,
|
| 384 |
+
"learning_rate": 1.842333954065798e-05,
|
| 385 |
+
"loss": 0.514,
|
| 386 |
+
"step": 1225
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 3.4926522043386985,
|
| 390 |
+
"grad_norm": 19.837966918945312,
|
| 391 |
+
"learning_rate": 1.834574798261949e-05,
|
| 392 |
+
"loss": 0.6475,
|
| 393 |
+
"step": 1250
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.562631210636809,
|
| 397 |
+
"grad_norm": 18.455398559570312,
|
| 398 |
+
"learning_rate": 1.8268156424581006e-05,
|
| 399 |
+
"loss": 0.6985,
|
| 400 |
+
"step": 1275
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.6326102169349195,
|
| 404 |
+
"grad_norm": 1.2877476215362549,
|
| 405 |
+
"learning_rate": 1.819056486654252e-05,
|
| 406 |
+
"loss": 0.7145,
|
| 407 |
+
"step": 1300
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.70258922323303,
|
| 411 |
+
"grad_norm": 46.33899688720703,
|
| 412 |
+
"learning_rate": 1.8112973308504036e-05,
|
| 413 |
+
"loss": 0.4402,
|
| 414 |
+
"step": 1325
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7725682295311405,
|
| 418 |
+
"grad_norm": 0.3435637652873993,
|
| 419 |
+
"learning_rate": 1.803538175046555e-05,
|
| 420 |
+
"loss": 0.5126,
|
| 421 |
+
"step": 1350
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.842547235829251,
|
| 425 |
+
"grad_norm": 1.4445362091064453,
|
| 426 |
+
"learning_rate": 1.7957790192427063e-05,
|
| 427 |
+
"loss": 0.5023,
|
| 428 |
+
"step": 1375
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.912526242127362,
|
| 432 |
+
"grad_norm": 0.9608930945396423,
|
| 433 |
+
"learning_rate": 1.788019863438858e-05,
|
| 434 |
+
"loss": 0.4957,
|
| 435 |
+
"step": 1400
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.9825052484254724,
|
| 439 |
+
"grad_norm": 10.732783317565918,
|
| 440 |
+
"learning_rate": 1.7802607076350093e-05,
|
| 441 |
+
"loss": 0.58,
|
| 442 |
+
"step": 1425
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0,
|
| 446 |
+
"eval_accuracy": 0.9469361970941251,
|
| 447 |
+
"eval_loss": 0.24154046177864075,
|
| 448 |
+
"eval_macro_f1": 0.8294877169586066,
|
| 449 |
+
"eval_runtime": 1.2972,
|
| 450 |
+
"eval_samples_per_second": 1220.359,
|
| 451 |
+
"eval_steps_per_second": 76.321,
|
| 452 |
+
"eval_weighted_f1": 0.9428227509673922,
|
| 453 |
+
"step": 1432
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 4.0503848845346395,
|
| 457 |
+
"grad_norm": 5.823998928070068,
|
| 458 |
+
"learning_rate": 1.772501551831161e-05,
|
| 459 |
+
"loss": 0.4628,
|
| 460 |
+
"step": 1450
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.12036389083275,
|
| 464 |
+
"grad_norm": 1.7952710390090942,
|
| 465 |
+
"learning_rate": 1.7647423960273124e-05,
|
| 466 |
+
"loss": 0.5368,
|
| 467 |
+
"step": 1475
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 4.1903428971308605,
|
| 471 |
+
"grad_norm": 1.0572470426559448,
|
| 472 |
+
"learning_rate": 1.756983240223464e-05,
|
| 473 |
+
"loss": 0.5197,
|
| 474 |
+
"step": 1500
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.260321903428971,
|
| 478 |
+
"grad_norm": 8.489173889160156,
|
| 479 |
+
"learning_rate": 1.7492240844196154e-05,
|
| 480 |
+
"loss": 0.6243,
|
| 481 |
+
"step": 1525
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.3303009097270815,
|
| 485 |
+
"grad_norm": 45.122440338134766,
|
| 486 |
+
"learning_rate": 1.741464928615767e-05,
|
| 487 |
+
"loss": 0.3062,
|
| 488 |
+
"step": 1550
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 4.400279916025193,
|
| 492 |
+
"grad_norm": 14.028529167175293,
|
| 493 |
+
"learning_rate": 1.733705772811918e-05,
|
| 494 |
+
"loss": 0.3844,
|
| 495 |
+
"step": 1575
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 4.470258922323303,
|
| 499 |
+
"grad_norm": 0.44512128829956055,
|
| 500 |
+
"learning_rate": 1.7259466170080696e-05,
|
| 501 |
+
"loss": 0.4883,
|
| 502 |
+
"step": 1600
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 4.540237928621414,
|
| 506 |
+
"grad_norm": 20.564664840698242,
|
| 507 |
+
"learning_rate": 1.718187461204221e-05,
|
| 508 |
+
"loss": 0.4162,
|
| 509 |
+
"step": 1625
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 4.610216934919524,
|
| 513 |
+
"grad_norm": 2.2896320819854736,
|
| 514 |
+
"learning_rate": 1.7104283054003726e-05,
|
| 515 |
+
"loss": 0.4814,
|
| 516 |
+
"step": 1650
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 4.680195941217635,
|
| 520 |
+
"grad_norm": 2.053428888320923,
|
| 521 |
+
"learning_rate": 1.702669149596524e-05,
|
| 522 |
+
"loss": 0.3705,
|
| 523 |
+
"step": 1675
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 4.750174947515745,
|
| 527 |
+
"grad_norm": 41.13652801513672,
|
| 528 |
+
"learning_rate": 1.6949099937926752e-05,
|
| 529 |
+
"loss": 0.4588,
|
| 530 |
+
"step": 1700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 4.820153953813856,
|
| 534 |
+
"grad_norm": 3.011770248413086,
|
| 535 |
+
"learning_rate": 1.687150837988827e-05,
|
| 536 |
+
"loss": 0.4399,
|
| 537 |
+
"step": 1725
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.890132960111966,
|
| 541 |
+
"grad_norm": 15.254034042358398,
|
| 542 |
+
"learning_rate": 1.6793916821849786e-05,
|
| 543 |
+
"loss": 0.247,
|
| 544 |
+
"step": 1750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.960111966410077,
|
| 548 |
+
"grad_norm": 14.806401252746582,
|
| 549 |
+
"learning_rate": 1.67163252638113e-05,
|
| 550 |
+
"loss": 0.5831,
|
| 551 |
+
"step": 1775
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 5.0,
|
| 555 |
+
"eval_accuracy": 0.9406190777005685,
|
| 556 |
+
"eval_loss": 0.26180794835090637,
|
| 557 |
+
"eval_macro_f1": 0.8366219283418204,
|
| 558 |
+
"eval_runtime": 1.2905,
|
| 559 |
+
"eval_samples_per_second": 1226.621,
|
| 560 |
+
"eval_steps_per_second": 76.712,
|
| 561 |
+
"eval_weighted_f1": 0.9416365846055232,
|
| 562 |
+
"step": 1790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 5.027991602519244,
|
| 566 |
+
"grad_norm": 49.11066436767578,
|
| 567 |
+
"learning_rate": 1.6638733705772813e-05,
|
| 568 |
+
"loss": 0.4847,
|
| 569 |
+
"step": 1800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 5.097970608817355,
|
| 573 |
+
"grad_norm": 3.138051748275757,
|
| 574 |
+
"learning_rate": 1.6561142147734328e-05,
|
| 575 |
+
"loss": 0.3709,
|
| 576 |
+
"step": 1825
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 5.167949615115465,
|
| 580 |
+
"grad_norm": 68.00996398925781,
|
| 581 |
+
"learning_rate": 1.6483550589695843e-05,
|
| 582 |
+
"loss": 0.4448,
|
| 583 |
+
"step": 1850
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 5.237928621413576,
|
| 587 |
+
"grad_norm": 0.19001732766628265,
|
| 588 |
+
"learning_rate": 1.6405959031657358e-05,
|
| 589 |
+
"loss": 0.2663,
|
| 590 |
+
"step": 1875
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 5.307907627711686,
|
| 594 |
+
"grad_norm": 0.25441697239875793,
|
| 595 |
+
"learning_rate": 1.632836747361887e-05,
|
| 596 |
+
"loss": 0.3225,
|
| 597 |
+
"step": 1900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 5.377886634009797,
|
| 601 |
+
"grad_norm": 17.98102378845215,
|
| 602 |
+
"learning_rate": 1.6250775915580385e-05,
|
| 603 |
+
"loss": 0.6131,
|
| 604 |
+
"step": 1925
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 5.447865640307907,
|
| 608 |
+
"grad_norm": 2.8782224655151367,
|
| 609 |
+
"learning_rate": 1.61731843575419e-05,
|
| 610 |
+
"loss": 0.4778,
|
| 611 |
+
"step": 1950
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 5.517844646606018,
|
| 615 |
+
"grad_norm": 48.173675537109375,
|
| 616 |
+
"learning_rate": 1.6095592799503415e-05,
|
| 617 |
+
"loss": 0.3094,
|
| 618 |
+
"step": 1975
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 5.587823652904129,
|
| 622 |
+
"grad_norm": 78.5799789428711,
|
| 623 |
+
"learning_rate": 1.601800124146493e-05,
|
| 624 |
+
"loss": 0.4037,
|
| 625 |
+
"step": 2000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 5.65780265920224,
|
| 629 |
+
"grad_norm": 0.39592620730400085,
|
| 630 |
+
"learning_rate": 1.5940409683426445e-05,
|
| 631 |
+
"loss": 0.3915,
|
| 632 |
+
"step": 2025
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 5.72778166550035,
|
| 636 |
+
"grad_norm": 0.23372972011566162,
|
| 637 |
+
"learning_rate": 1.586281812538796e-05,
|
| 638 |
+
"loss": 0.3545,
|
| 639 |
+
"step": 2050
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.797760671798461,
|
| 643 |
+
"grad_norm": 0.20506474375724792,
|
| 644 |
+
"learning_rate": 1.5785226567349475e-05,
|
| 645 |
+
"loss": 0.3706,
|
| 646 |
+
"step": 2075
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.867739678096571,
|
| 650 |
+
"grad_norm": 7.170314311981201,
|
| 651 |
+
"learning_rate": 1.570763500931099e-05,
|
| 652 |
+
"loss": 0.2673,
|
| 653 |
+
"step": 2100
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.937718684394682,
|
| 657 |
+
"grad_norm": 0.43047407269477844,
|
| 658 |
+
"learning_rate": 1.5630043451272502e-05,
|
| 659 |
+
"loss": 0.4991,
|
| 660 |
+
"step": 2125
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.0,
|
| 664 |
+
"eval_accuracy": 0.9538850284270373,
|
| 665 |
+
"eval_loss": 0.24408170580863953,
|
| 666 |
+
"eval_macro_f1": 0.8642730226359241,
|
| 667 |
+
"eval_runtime": 1.2865,
|
| 668 |
+
"eval_samples_per_second": 1230.458,
|
| 669 |
+
"eval_steps_per_second": 76.952,
|
| 670 |
+
"eval_weighted_f1": 0.9517727315836694,
|
| 671 |
+
"step": 2148
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 6.005598320503849,
|
| 675 |
+
"grad_norm": 26.87923812866211,
|
| 676 |
+
"learning_rate": 1.5552451893234017e-05,
|
| 677 |
+
"loss": 0.2136,
|
| 678 |
+
"step": 2150
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 6.07557732680196,
|
| 682 |
+
"grad_norm": 200.5290985107422,
|
| 683 |
+
"learning_rate": 1.5474860335195532e-05,
|
| 684 |
+
"loss": 0.2688,
|
| 685 |
+
"step": 2175
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 6.14555633310007,
|
| 689 |
+
"grad_norm": 0.23392556607723236,
|
| 690 |
+
"learning_rate": 1.5397268777157047e-05,
|
| 691 |
+
"loss": 0.4127,
|
| 692 |
+
"step": 2200
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 6.215535339398181,
|
| 696 |
+
"grad_norm": 2.0901408195495605,
|
| 697 |
+
"learning_rate": 1.531967721911856e-05,
|
| 698 |
+
"loss": 0.3672,
|
| 699 |
+
"step": 2225
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 6.285514345696291,
|
| 703 |
+
"grad_norm": 85.51931762695312,
|
| 704 |
+
"learning_rate": 1.5242085661080076e-05,
|
| 705 |
+
"loss": 0.3108,
|
| 706 |
+
"step": 2250
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 6.355493351994402,
|
| 710 |
+
"grad_norm": 120.88185119628906,
|
| 711 |
+
"learning_rate": 1.5164494103041589e-05,
|
| 712 |
+
"loss": 0.4208,
|
| 713 |
+
"step": 2275
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.425472358292512,
|
| 717 |
+
"grad_norm": 0.24016176164150238,
|
| 718 |
+
"learning_rate": 1.5086902545003106e-05,
|
| 719 |
+
"loss": 0.2161,
|
| 720 |
+
"step": 2300
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.495451364590623,
|
| 724 |
+
"grad_norm": 10.961566925048828,
|
| 725 |
+
"learning_rate": 1.5009310986964621e-05,
|
| 726 |
+
"loss": 0.2972,
|
| 727 |
+
"step": 2325
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 6.565430370888733,
|
| 731 |
+
"grad_norm": 23.026823043823242,
|
| 732 |
+
"learning_rate": 1.4931719428926134e-05,
|
| 733 |
+
"loss": 0.2287,
|
| 734 |
+
"step": 2350
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 6.635409377186844,
|
| 738 |
+
"grad_norm": 71.0625,
|
| 739 |
+
"learning_rate": 1.485412787088765e-05,
|
| 740 |
+
"loss": 0.3265,
|
| 741 |
+
"step": 2375
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 6.705388383484954,
|
| 745 |
+
"grad_norm": 119.8893814086914,
|
| 746 |
+
"learning_rate": 1.4776536312849163e-05,
|
| 747 |
+
"loss": 0.396,
|
| 748 |
+
"step": 2400
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 6.7753673897830655,
|
| 752 |
+
"grad_norm": 34.25016784667969,
|
| 753 |
+
"learning_rate": 1.4698944754810678e-05,
|
| 754 |
+
"loss": 0.4438,
|
| 755 |
+
"step": 2425
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 6.845346396081176,
|
| 759 |
+
"grad_norm": 0.3095424771308899,
|
| 760 |
+
"learning_rate": 1.4621353196772193e-05,
|
| 761 |
+
"loss": 0.3035,
|
| 762 |
+
"step": 2450
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 6.9153254023792865,
|
| 766 |
+
"grad_norm": 1.8962700366973877,
|
| 767 |
+
"learning_rate": 1.4543761638733706e-05,
|
| 768 |
+
"loss": 0.2966,
|
| 769 |
+
"step": 2475
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 6.985304408677397,
|
| 773 |
+
"grad_norm": 18.388277053833008,
|
| 774 |
+
"learning_rate": 1.4466170080695221e-05,
|
| 775 |
+
"loss": 0.2788,
|
| 776 |
+
"step": 2500
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 7.0,
|
| 780 |
+
"eval_accuracy": 0.9513581806696146,
|
| 781 |
+
"eval_loss": 0.2554718554019928,
|
| 782 |
+
"eval_macro_f1": 0.8662234550886146,
|
| 783 |
+
"eval_runtime": 1.3015,
|
| 784 |
+
"eval_samples_per_second": 1216.247,
|
| 785 |
+
"eval_steps_per_second": 76.063,
|
| 786 |
+
"eval_weighted_f1": 0.9503012319701283,
|
| 787 |
+
"step": 2506
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 7.053184044786564,
|
| 791 |
+
"grad_norm": 0.6708895564079285,
|
| 792 |
+
"learning_rate": 1.4388578522656735e-05,
|
| 793 |
+
"loss": 0.3312,
|
| 794 |
+
"step": 2525
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 7.123163051084674,
|
| 798 |
+
"grad_norm": 0.1899476796388626,
|
| 799 |
+
"learning_rate": 1.431098696461825e-05,
|
| 800 |
+
"loss": 0.3434,
|
| 801 |
+
"step": 2550
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 7.193142057382786,
|
| 805 |
+
"grad_norm": 0.3260430693626404,
|
| 806 |
+
"learning_rate": 1.4233395406579767e-05,
|
| 807 |
+
"loss": 0.4349,
|
| 808 |
+
"step": 2575
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 7.263121063680896,
|
| 812 |
+
"grad_norm": 2.4518017768859863,
|
| 813 |
+
"learning_rate": 1.415580384854128e-05,
|
| 814 |
+
"loss": 0.1875,
|
| 815 |
+
"step": 2600
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 7.3331000699790065,
|
| 819 |
+
"grad_norm": 17.725204467773438,
|
| 820 |
+
"learning_rate": 1.4078212290502795e-05,
|
| 821 |
+
"loss": 0.2664,
|
| 822 |
+
"step": 2625
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 7.403079076277117,
|
| 826 |
+
"grad_norm": 0.14622612297534943,
|
| 827 |
+
"learning_rate": 1.400062073246431e-05,
|
| 828 |
+
"loss": 0.1896,
|
| 829 |
+
"step": 2650
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 7.4730580825752275,
|
| 833 |
+
"grad_norm": 0.13795089721679688,
|
| 834 |
+
"learning_rate": 1.3923029174425823e-05,
|
| 835 |
+
"loss": 0.3782,
|
| 836 |
+
"step": 2675
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.543037088873338,
|
| 840 |
+
"grad_norm": 0.27982577681541443,
|
| 841 |
+
"learning_rate": 1.3845437616387339e-05,
|
| 842 |
+
"loss": 0.2526,
|
| 843 |
+
"step": 2700
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.6130160951714485,
|
| 847 |
+
"grad_norm": 36.00263595581055,
|
| 848 |
+
"learning_rate": 1.3767846058348852e-05,
|
| 849 |
+
"loss": 0.2975,
|
| 850 |
+
"step": 2725
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.682995101469559,
|
| 854 |
+
"grad_norm": 1.6170415878295898,
|
| 855 |
+
"learning_rate": 1.3690254500310367e-05,
|
| 856 |
+
"loss": 0.2923,
|
| 857 |
+
"step": 2750
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.7529741077676695,
|
| 861 |
+
"grad_norm": 0.2949804961681366,
|
| 862 |
+
"learning_rate": 1.3612662942271882e-05,
|
| 863 |
+
"loss": 0.3919,
|
| 864 |
+
"step": 2775
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.82295311406578,
|
| 868 |
+
"grad_norm": 0.23040179908275604,
|
| 869 |
+
"learning_rate": 1.3535071384233395e-05,
|
| 870 |
+
"loss": 0.2382,
|
| 871 |
+
"step": 2800
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.8929321203638905,
|
| 875 |
+
"grad_norm": 21.18236541748047,
|
| 876 |
+
"learning_rate": 1.345747982619491e-05,
|
| 877 |
+
"loss": 0.2501,
|
| 878 |
+
"step": 2825
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.962911126662002,
|
| 882 |
+
"grad_norm": 19.844017028808594,
|
| 883 |
+
"learning_rate": 1.3379888268156424e-05,
|
| 884 |
+
"loss": 0.3657,
|
| 885 |
+
"step": 2850
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 8.0,
|
| 889 |
+
"eval_accuracy": 0.9519898926089703,
|
| 890 |
+
"eval_loss": 0.26367440819740295,
|
| 891 |
+
"eval_macro_f1": 0.8560865023759766,
|
| 892 |
+
"eval_runtime": 1.2858,
|
| 893 |
+
"eval_samples_per_second": 1231.145,
|
| 894 |
+
"eval_steps_per_second": 76.995,
|
| 895 |
+
"eval_weighted_f1": 0.9495265878908666,
|
| 896 |
+
"step": 2864
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 8.030790762771169,
|
| 900 |
+
"grad_norm": 0.19330187141895294,
|
| 901 |
+
"learning_rate": 1.330229671011794e-05,
|
| 902 |
+
"loss": 0.2496,
|
| 903 |
+
"step": 2875
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 8.100769769069279,
|
| 907 |
+
"grad_norm": 0.13222865760326385,
|
| 908 |
+
"learning_rate": 1.3224705152079456e-05,
|
| 909 |
+
"loss": 0.2939,
|
| 910 |
+
"step": 2900
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 8.17074877536739,
|
| 914 |
+
"grad_norm": 0.1420418620109558,
|
| 915 |
+
"learning_rate": 1.314711359404097e-05,
|
| 916 |
+
"loss": 0.1916,
|
| 917 |
+
"step": 2925
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 8.2407277816655,
|
| 921 |
+
"grad_norm": 79.88240814208984,
|
| 922 |
+
"learning_rate": 1.3069522036002484e-05,
|
| 923 |
+
"loss": 0.2371,
|
| 924 |
+
"step": 2950
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 8.31070678796361,
|
| 928 |
+
"grad_norm": 1.5356426239013672,
|
| 929 |
+
"learning_rate": 1.2991930477964e-05,
|
| 930 |
+
"loss": 0.2095,
|
| 931 |
+
"step": 2975
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 8.380685794261721,
|
| 935 |
+
"grad_norm": 36.07261276245117,
|
| 936 |
+
"learning_rate": 1.2914338919925513e-05,
|
| 937 |
+
"loss": 0.2685,
|
| 938 |
+
"step": 3000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 8.450664800559831,
|
| 942 |
+
"grad_norm": 2.8488714694976807,
|
| 943 |
+
"learning_rate": 1.2836747361887028e-05,
|
| 944 |
+
"loss": 0.2696,
|
| 945 |
+
"step": 3025
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 8.520643806857942,
|
| 949 |
+
"grad_norm": 113.312744140625,
|
| 950 |
+
"learning_rate": 1.2759155803848541e-05,
|
| 951 |
+
"loss": 0.2683,
|
| 952 |
+
"step": 3050
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 8.590622813156052,
|
| 956 |
+
"grad_norm": 148.375,
|
| 957 |
+
"learning_rate": 1.2681564245810056e-05,
|
| 958 |
+
"loss": 0.1597,
|
| 959 |
+
"step": 3075
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 8.660601819454163,
|
| 963 |
+
"grad_norm": 0.23184065520763397,
|
| 964 |
+
"learning_rate": 1.2603972687771571e-05,
|
| 965 |
+
"loss": 0.126,
|
| 966 |
+
"step": 3100
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 8.730580825752273,
|
| 970 |
+
"grad_norm": 3.8853540420532227,
|
| 971 |
+
"learning_rate": 1.2526381129733085e-05,
|
| 972 |
+
"loss": 0.2859,
|
| 973 |
+
"step": 3125
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 8.800559832050386,
|
| 977 |
+
"grad_norm": 2.980635166168213,
|
| 978 |
+
"learning_rate": 1.2448789571694601e-05,
|
| 979 |
+
"loss": 0.4045,
|
| 980 |
+
"step": 3150
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 8.870538838348496,
|
| 984 |
+
"grad_norm": 24.395973205566406,
|
| 985 |
+
"learning_rate": 1.2371198013656117e-05,
|
| 986 |
+
"loss": 0.2076,
|
| 987 |
+
"step": 3175
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 8.940517844646607,
|
| 991 |
+
"grad_norm": 0.11534596234560013,
|
| 992 |
+
"learning_rate": 1.229360645561763e-05,
|
| 993 |
+
"loss": 0.1153,
|
| 994 |
+
"step": 3200
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 9.0,
|
| 998 |
+
"eval_accuracy": 0.9494630448515476,
|
| 999 |
+
"eval_loss": 0.275835782289505,
|
| 1000 |
+
"eval_macro_f1": 0.8595203810510755,
|
| 1001 |
+
"eval_runtime": 1.2825,
|
| 1002 |
+
"eval_samples_per_second": 1234.339,
|
| 1003 |
+
"eval_steps_per_second": 77.195,
|
| 1004 |
+
"eval_weighted_f1": 0.9490024441092851,
|
| 1005 |
+
"step": 3222
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 9.008397480755773,
|
| 1009 |
+
"grad_norm": 0.11499301344156265,
|
| 1010 |
+
"learning_rate": 1.2216014897579145e-05,
|
| 1011 |
+
"loss": 0.3014,
|
| 1012 |
+
"step": 3225
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 9.078376487053884,
|
| 1016 |
+
"grad_norm": 3.1832170486450195,
|
| 1017 |
+
"learning_rate": 1.213842333954066e-05,
|
| 1018 |
+
"loss": 0.3592,
|
| 1019 |
+
"step": 3250
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 9.148355493351994,
|
| 1023 |
+
"grad_norm": 70.31085968017578,
|
| 1024 |
+
"learning_rate": 1.2060831781502173e-05,
|
| 1025 |
+
"loss": 0.1704,
|
| 1026 |
+
"step": 3275
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 9.218334499650105,
|
| 1030 |
+
"grad_norm": 0.11831537634134293,
|
| 1031 |
+
"learning_rate": 1.1983240223463688e-05,
|
| 1032 |
+
"loss": 0.1298,
|
| 1033 |
+
"step": 3300
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 9.288313505948215,
|
| 1037 |
+
"grad_norm": 0.16945354640483856,
|
| 1038 |
+
"learning_rate": 1.1905648665425202e-05,
|
| 1039 |
+
"loss": 0.2819,
|
| 1040 |
+
"step": 3325
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 9.358292512246326,
|
| 1044 |
+
"grad_norm": 3.1526803970336914,
|
| 1045 |
+
"learning_rate": 1.1828057107386717e-05,
|
| 1046 |
+
"loss": 0.131,
|
| 1047 |
+
"step": 3350
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 9.428271518544436,
|
| 1051 |
+
"grad_norm": 66.15129852294922,
|
| 1052 |
+
"learning_rate": 1.175046554934823e-05,
|
| 1053 |
+
"loss": 0.2481,
|
| 1054 |
+
"step": 3375
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 9.498250524842547,
|
| 1058 |
+
"grad_norm": 0.18026110529899597,
|
| 1059 |
+
"learning_rate": 1.1672873991309745e-05,
|
| 1060 |
+
"loss": 0.1182,
|
| 1061 |
+
"step": 3400
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 9.568229531140657,
|
| 1065 |
+
"grad_norm": 4.617343902587891,
|
| 1066 |
+
"learning_rate": 1.1595282433271262e-05,
|
| 1067 |
+
"loss": 0.1801,
|
| 1068 |
+
"step": 3425
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 9.638208537438768,
|
| 1072 |
+
"grad_norm": 0.13688743114471436,
|
| 1073 |
+
"learning_rate": 1.1517690875232777e-05,
|
| 1074 |
+
"loss": 0.1429,
|
| 1075 |
+
"step": 3450
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 9.708187543736878,
|
| 1079 |
+
"grad_norm": 0.22725597023963928,
|
| 1080 |
+
"learning_rate": 1.144009931719429e-05,
|
| 1081 |
+
"loss": 0.2457,
|
| 1082 |
+
"step": 3475
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 9.778166550034989,
|
| 1086 |
+
"grad_norm": 4.924619674682617,
|
| 1087 |
+
"learning_rate": 1.1362507759155806e-05,
|
| 1088 |
+
"loss": 0.1405,
|
| 1089 |
+
"step": 3500
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 9.8481455563331,
|
| 1093 |
+
"grad_norm": 0.08874160051345825,
|
| 1094 |
+
"learning_rate": 1.1284916201117319e-05,
|
| 1095 |
+
"loss": 0.1614,
|
| 1096 |
+
"step": 3525
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 9.91812456263121,
|
| 1100 |
+
"grad_norm": 21.604572296142578,
|
| 1101 |
+
"learning_rate": 1.1207324643078834e-05,
|
| 1102 |
+
"loss": 0.3198,
|
| 1103 |
+
"step": 3550
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 9.988103568929322,
|
| 1107 |
+
"grad_norm": 0.0932706743478775,
|
| 1108 |
+
"learning_rate": 1.112973308504035e-05,
|
| 1109 |
+
"loss": 0.1274,
|
| 1110 |
+
"step": 3575
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 10.0,
|
| 1114 |
+
"eval_accuracy": 0.9456727732154138,
|
| 1115 |
+
"eval_loss": 0.30851510167121887,
|
| 1116 |
+
"eval_macro_f1": 0.8524275587082389,
|
| 1117 |
+
"eval_runtime": 1.2805,
|
| 1118 |
+
"eval_samples_per_second": 1236.21,
|
| 1119 |
+
"eval_steps_per_second": 77.312,
|
| 1120 |
+
"eval_weighted_f1": 0.9438806903485606,
|
| 1121 |
+
"step": 3580
|
| 1122 |
+
}
|
| 1123 |
+
],
|
| 1124 |
+
"logging_steps": 25,
|
| 1125 |
+
"max_steps": 7160,
|
| 1126 |
+
"num_input_tokens_seen": 0,
|
| 1127 |
+
"num_train_epochs": 20,
|
| 1128 |
+
"save_steps": 500,
|
| 1129 |
+
"stateful_callbacks": {
|
| 1130 |
+
"EarlyStoppingCallback": {
|
| 1131 |
+
"args": {
|
| 1132 |
+
"early_stopping_patience": 3,
|
| 1133 |
+
"early_stopping_threshold": 0.0
|
| 1134 |
+
},
|
| 1135 |
+
"attributes": {
|
| 1136 |
+
"early_stopping_patience_counter": 3
|
| 1137 |
+
}
|
| 1138 |
+
},
|
| 1139 |
+
"TrainerControl": {
|
| 1140 |
+
"args": {
|
| 1141 |
+
"should_epoch_stop": false,
|
| 1142 |
+
"should_evaluate": false,
|
| 1143 |
+
"should_log": false,
|
| 1144 |
+
"should_save": true,
|
| 1145 |
+
"should_training_stop": true
|
| 1146 |
+
},
|
| 1147 |
+
"attributes": {}
|
| 1148 |
+
}
|
| 1149 |
+
},
|
| 1150 |
+
"total_flos": 0.0,
|
| 1151 |
+
"train_batch_size": 8,
|
| 1152 |
+
"trial_name": null,
|
| 1153 |
+
"trial_params": null
|
| 1154 |
+
}
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1d15437a12ab74f1b55f89335b5998bd3326c84823b685185c3f07be37b95cb
|
| 3 |
+
size 5560
|
uit_vsfc-phobert-s44-20260811T120119Z-6beb231d/trainer/checkpoint-3580/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fa35ce1aa34e87773b4d9242153661f79055fe9879fb26cc31ee0e0507ca3a1
|
| 3 |
+
size 540023320
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<mask>": 64000
|
| 3 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/bpe.codes
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e443183ccd1f59a6f2ea2ceb03387c3df8781aa874ee26758675eeb9e5b5e1d
|
| 3 |
+
size 540023352
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7c71a2d4dab2775cdd6de1c51beed39be70460fe33bb9fc0c41203ad65f2a2b
|
| 3 |
+
size 1075440186
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e65ab16deb8f37d3446f585f9f65d4fbf5a9a67327658a7f15e3cb2ba6c2aa5
|
| 3 |
+
size 14244
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ee49c0fe101000e4a7d6c06dd65ade764fad14d801bbd4e3800ecdc524dcb1d
|
| 3 |
+
size 1064
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/special_tokens_map.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": "<mask>",
|
| 6 |
+
"pad_token": "<pad>",
|
| 7 |
+
"sep_token": "</s>",
|
| 8 |
+
"unk_token": "<unk>"
|
| 9 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"64000": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 256,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "PhobertTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/trainer_state.json
ADDED
|
@@ -0,0 +1,139 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 225,
|
| 3 |
+
"best_metric": 0.8316730322479033,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225",
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 225,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.33444816053511706,
|
| 14 |
+
"grad_norm": 9.175091743469238,
|
| 15 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 16 |
+
"loss": 2.5402,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.6688963210702341,
|
| 21 |
+
"grad_norm": 8.813733100891113,
|
| 22 |
+
"learning_rate": 6.533333333333334e-06,
|
| 23 |
+
"loss": 2.4521,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"grad_norm": 7.533299922943115,
|
| 29 |
+
"learning_rate": 9.866666666666668e-06,
|
| 30 |
+
"loss": 2.1914,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.0,
|
| 35 |
+
"eval_accuracy": 0.783625730994152,
|
| 36 |
+
"eval_loss": 0.49943748116493225,
|
| 37 |
+
"eval_macro_f1": 0.6899289389855428,
|
| 38 |
+
"eval_runtime": 0.4173,
|
| 39 |
+
"eval_samples_per_second": 819.5,
|
| 40 |
+
"eval_steps_per_second": 52.716,
|
| 41 |
+
"eval_weighted_f1": 0.7537224995020427,
|
| 42 |
+
"step": 75
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.334448160535117,
|
| 46 |
+
"grad_norm": 23.902313232421875,
|
| 47 |
+
"learning_rate": 1.3200000000000002e-05,
|
| 48 |
+
"loss": 1.7157,
|
| 49 |
+
"step": 100
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.6688963210702341,
|
| 53 |
+
"grad_norm": 17.904226303100586,
|
| 54 |
+
"learning_rate": 1.6533333333333333e-05,
|
| 55 |
+
"loss": 1.6271,
|
| 56 |
+
"step": 125
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 2.0,
|
| 60 |
+
"grad_norm": 16.19369888305664,
|
| 61 |
+
"learning_rate": 1.9866666666666667e-05,
|
| 62 |
+
"loss": 1.4274,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 2.0,
|
| 67 |
+
"eval_accuracy": 0.8391812865497076,
|
| 68 |
+
"eval_loss": 0.34102874994277954,
|
| 69 |
+
"eval_macro_f1": 0.813457628799524,
|
| 70 |
+
"eval_runtime": 0.3999,
|
| 71 |
+
"eval_samples_per_second": 855.277,
|
| 72 |
+
"eval_steps_per_second": 55.018,
|
| 73 |
+
"eval_weighted_f1": 0.8393838350359296,
|
| 74 |
+
"step": 150
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.334448160535117,
|
| 78 |
+
"grad_norm": 39.397132873535156,
|
| 79 |
+
"learning_rate": 1.9644444444444447e-05,
|
| 80 |
+
"loss": 0.9702,
|
| 81 |
+
"step": 175
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.668896321070234,
|
| 85 |
+
"grad_norm": 20.064456939697266,
|
| 86 |
+
"learning_rate": 1.9274074074074076e-05,
|
| 87 |
+
"loss": 1.1357,
|
| 88 |
+
"step": 200
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 3.0,
|
| 92 |
+
"grad_norm": 26.795101165771484,
|
| 93 |
+
"learning_rate": 1.8903703703703705e-05,
|
| 94 |
+
"loss": 1.1531,
|
| 95 |
+
"step": 225
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.0,
|
| 99 |
+
"eval_accuracy": 0.8538011695906432,
|
| 100 |
+
"eval_loss": 0.35740259289741516,
|
| 101 |
+
"eval_macro_f1": 0.8316730322479033,
|
| 102 |
+
"eval_runtime": 0.4053,
|
| 103 |
+
"eval_samples_per_second": 843.731,
|
| 104 |
+
"eval_steps_per_second": 54.275,
|
| 105 |
+
"eval_weighted_f1": 0.8545149804726672,
|
| 106 |
+
"step": 225
|
| 107 |
+
}
|
| 108 |
+
],
|
| 109 |
+
"logging_steps": 25,
|
| 110 |
+
"max_steps": 1500,
|
| 111 |
+
"num_input_tokens_seen": 0,
|
| 112 |
+
"num_train_epochs": 20,
|
| 113 |
+
"save_steps": 500,
|
| 114 |
+
"stateful_callbacks": {
|
| 115 |
+
"EarlyStoppingCallback": {
|
| 116 |
+
"args": {
|
| 117 |
+
"early_stopping_patience": 3,
|
| 118 |
+
"early_stopping_threshold": 0.0
|
| 119 |
+
},
|
| 120 |
+
"attributes": {
|
| 121 |
+
"early_stopping_patience_counter": 0
|
| 122 |
+
}
|
| 123 |
+
},
|
| 124 |
+
"TrainerControl": {
|
| 125 |
+
"args": {
|
| 126 |
+
"should_epoch_stop": false,
|
| 127 |
+
"should_evaluate": false,
|
| 128 |
+
"should_log": false,
|
| 129 |
+
"should_save": true,
|
| 130 |
+
"should_training_stop": false
|
| 131 |
+
},
|
| 132 |
+
"attributes": {}
|
| 133 |
+
}
|
| 134 |
+
},
|
| 135 |
+
"total_flos": 0.0,
|
| 136 |
+
"train_batch_size": 8,
|
| 137 |
+
"trial_name": null,
|
| 138 |
+
"trial_params": null
|
| 139 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:349135d230d1ed107ab68d23c12e1ec29e33ead7aa6bb5d9825a51628b6a7642
|
| 3 |
+
size 5560
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<mask>": 64000
|
| 3 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/bpe.codes
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2081e5153aec566ea01e32585fd4b2c9e46fe548cfd2f650cbb903161097925
|
| 3 |
+
size 540023352
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54f22df87fd8ed33bfe9e5bf161325254512e83da936974416bd218d0b1010cc
|
| 3 |
+
size 1075440186
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ff5772ecc3aaaca5f14a63bef1cc9a8de4120d07caf3c4bc6f92fe667da6e73
|
| 3 |
+
size 14244
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0821cfdf96ff4f5fe5e0d5bd5c895e1383e459df8e6e1d261b6d96bc93f4abc
|
| 3 |
+
size 1064
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/special_tokens_map.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<s>",
|
| 3 |
+
"cls_token": "<s>",
|
| 4 |
+
"eos_token": "</s>",
|
| 5 |
+
"mask_token": "<mask>",
|
| 6 |
+
"pad_token": "<pad>",
|
| 7 |
+
"sep_token": "</s>",
|
| 8 |
+
"unk_token": "<unk>"
|
| 9 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/tokenizer_config.json
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<pad>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "<unk>",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": false,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"64000": {
|
| 36 |
+
"content": "<mask>",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "<s>",
|
| 45 |
+
"clean_up_tokenization_spaces": false,
|
| 46 |
+
"cls_token": "<s>",
|
| 47 |
+
"eos_token": "</s>",
|
| 48 |
+
"extra_special_tokens": {},
|
| 49 |
+
"mask_token": "<mask>",
|
| 50 |
+
"model_max_length": 256,
|
| 51 |
+
"pad_token": "<pad>",
|
| 52 |
+
"sep_token": "</s>",
|
| 53 |
+
"tokenizer_class": "PhobertTokenizer",
|
| 54 |
+
"unk_token": "<unk>"
|
| 55 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/trainer_state.json
ADDED
|
@@ -0,0 +1,235 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 225,
|
| 3 |
+
"best_metric": 0.8316730322479033,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-225",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 450,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.33444816053511706,
|
| 14 |
+
"grad_norm": 9.175091743469238,
|
| 15 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 16 |
+
"loss": 2.5402,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.6688963210702341,
|
| 21 |
+
"grad_norm": 8.813733100891113,
|
| 22 |
+
"learning_rate": 6.533333333333334e-06,
|
| 23 |
+
"loss": 2.4521,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.0,
|
| 28 |
+
"grad_norm": 7.533299922943115,
|
| 29 |
+
"learning_rate": 9.866666666666668e-06,
|
| 30 |
+
"loss": 2.1914,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 1.0,
|
| 35 |
+
"eval_accuracy": 0.783625730994152,
|
| 36 |
+
"eval_loss": 0.49943748116493225,
|
| 37 |
+
"eval_macro_f1": 0.6899289389855428,
|
| 38 |
+
"eval_runtime": 0.4173,
|
| 39 |
+
"eval_samples_per_second": 819.5,
|
| 40 |
+
"eval_steps_per_second": 52.716,
|
| 41 |
+
"eval_weighted_f1": 0.7537224995020427,
|
| 42 |
+
"step": 75
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 1.334448160535117,
|
| 46 |
+
"grad_norm": 23.902313232421875,
|
| 47 |
+
"learning_rate": 1.3200000000000002e-05,
|
| 48 |
+
"loss": 1.7157,
|
| 49 |
+
"step": 100
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.6688963210702341,
|
| 53 |
+
"grad_norm": 17.904226303100586,
|
| 54 |
+
"learning_rate": 1.6533333333333333e-05,
|
| 55 |
+
"loss": 1.6271,
|
| 56 |
+
"step": 125
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 2.0,
|
| 60 |
+
"grad_norm": 16.19369888305664,
|
| 61 |
+
"learning_rate": 1.9866666666666667e-05,
|
| 62 |
+
"loss": 1.4274,
|
| 63 |
+
"step": 150
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 2.0,
|
| 67 |
+
"eval_accuracy": 0.8391812865497076,
|
| 68 |
+
"eval_loss": 0.34102874994277954,
|
| 69 |
+
"eval_macro_f1": 0.813457628799524,
|
| 70 |
+
"eval_runtime": 0.3999,
|
| 71 |
+
"eval_samples_per_second": 855.277,
|
| 72 |
+
"eval_steps_per_second": 55.018,
|
| 73 |
+
"eval_weighted_f1": 0.8393838350359296,
|
| 74 |
+
"step": 150
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 2.334448160535117,
|
| 78 |
+
"grad_norm": 39.397132873535156,
|
| 79 |
+
"learning_rate": 1.9644444444444447e-05,
|
| 80 |
+
"loss": 0.9702,
|
| 81 |
+
"step": 175
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 2.668896321070234,
|
| 85 |
+
"grad_norm": 20.064456939697266,
|
| 86 |
+
"learning_rate": 1.9274074074074076e-05,
|
| 87 |
+
"loss": 1.1357,
|
| 88 |
+
"step": 200
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 3.0,
|
| 92 |
+
"grad_norm": 26.795101165771484,
|
| 93 |
+
"learning_rate": 1.8903703703703705e-05,
|
| 94 |
+
"loss": 1.1531,
|
| 95 |
+
"step": 225
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.0,
|
| 99 |
+
"eval_accuracy": 0.8538011695906432,
|
| 100 |
+
"eval_loss": 0.35740259289741516,
|
| 101 |
+
"eval_macro_f1": 0.8316730322479033,
|
| 102 |
+
"eval_runtime": 0.4053,
|
| 103 |
+
"eval_samples_per_second": 843.731,
|
| 104 |
+
"eval_steps_per_second": 54.275,
|
| 105 |
+
"eval_weighted_f1": 0.8545149804726672,
|
| 106 |
+
"step": 225
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 3.334448160535117,
|
| 110 |
+
"grad_norm": 53.496341705322266,
|
| 111 |
+
"learning_rate": 1.8533333333333334e-05,
|
| 112 |
+
"loss": 0.7746,
|
| 113 |
+
"step": 250
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 3.668896321070234,
|
| 117 |
+
"grad_norm": 22.859004974365234,
|
| 118 |
+
"learning_rate": 1.8162962962962963e-05,
|
| 119 |
+
"loss": 0.7344,
|
| 120 |
+
"step": 275
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 4.0,
|
| 124 |
+
"grad_norm": 8.84278678894043,
|
| 125 |
+
"learning_rate": 1.7792592592592596e-05,
|
| 126 |
+
"loss": 0.659,
|
| 127 |
+
"step": 300
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 4.0,
|
| 131 |
+
"eval_accuracy": 0.847953216374269,
|
| 132 |
+
"eval_loss": 0.4698496162891388,
|
| 133 |
+
"eval_macro_f1": 0.8304089105889534,
|
| 134 |
+
"eval_runtime": 0.4233,
|
| 135 |
+
"eval_samples_per_second": 807.844,
|
| 136 |
+
"eval_steps_per_second": 51.967,
|
| 137 |
+
"eval_weighted_f1": 0.8508241027755025,
|
| 138 |
+
"step": 300
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 4.334448160535117,
|
| 142 |
+
"grad_norm": 16.639404296875,
|
| 143 |
+
"learning_rate": 1.7422222222222222e-05,
|
| 144 |
+
"loss": 0.3661,
|
| 145 |
+
"step": 325
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 4.668896321070234,
|
| 149 |
+
"grad_norm": 60.14759063720703,
|
| 150 |
+
"learning_rate": 1.7051851851851854e-05,
|
| 151 |
+
"loss": 0.408,
|
| 152 |
+
"step": 350
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 5.0,
|
| 156 |
+
"grad_norm": 24.689102172851562,
|
| 157 |
+
"learning_rate": 1.6681481481481484e-05,
|
| 158 |
+
"loss": 0.6895,
|
| 159 |
+
"step": 375
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 5.0,
|
| 163 |
+
"eval_accuracy": 0.847953216374269,
|
| 164 |
+
"eval_loss": 0.4754910469055176,
|
| 165 |
+
"eval_macro_f1": 0.8257836990595612,
|
| 166 |
+
"eval_runtime": 0.41,
|
| 167 |
+
"eval_samples_per_second": 834.076,
|
| 168 |
+
"eval_steps_per_second": 53.654,
|
| 169 |
+
"eval_weighted_f1": 0.8490435205045006,
|
| 170 |
+
"step": 375
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.334448160535117,
|
| 174 |
+
"grad_norm": 154.72659301757812,
|
| 175 |
+
"learning_rate": 1.6311111111111113e-05,
|
| 176 |
+
"loss": 0.2603,
|
| 177 |
+
"step": 400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 5.668896321070234,
|
| 181 |
+
"grad_norm": 17.728374481201172,
|
| 182 |
+
"learning_rate": 1.5940740740740742e-05,
|
| 183 |
+
"loss": 0.3168,
|
| 184 |
+
"step": 425
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 6.0,
|
| 188 |
+
"grad_norm": 26.6679744720459,
|
| 189 |
+
"learning_rate": 1.557037037037037e-05,
|
| 190 |
+
"loss": 0.5243,
|
| 191 |
+
"step": 450
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 6.0,
|
| 195 |
+
"eval_accuracy": 0.8333333333333334,
|
| 196 |
+
"eval_loss": 0.6150423288345337,
|
| 197 |
+
"eval_macro_f1": 0.7929760096853329,
|
| 198 |
+
"eval_runtime": 0.4069,
|
| 199 |
+
"eval_samples_per_second": 840.422,
|
| 200 |
+
"eval_steps_per_second": 54.062,
|
| 201 |
+
"eval_weighted_f1": 0.827186191320724,
|
| 202 |
+
"step": 450
|
| 203 |
+
}
|
| 204 |
+
],
|
| 205 |
+
"logging_steps": 25,
|
| 206 |
+
"max_steps": 1500,
|
| 207 |
+
"num_input_tokens_seen": 0,
|
| 208 |
+
"num_train_epochs": 20,
|
| 209 |
+
"save_steps": 500,
|
| 210 |
+
"stateful_callbacks": {
|
| 211 |
+
"EarlyStoppingCallback": {
|
| 212 |
+
"args": {
|
| 213 |
+
"early_stopping_patience": 3,
|
| 214 |
+
"early_stopping_threshold": 0.0
|
| 215 |
+
},
|
| 216 |
+
"attributes": {
|
| 217 |
+
"early_stopping_patience_counter": 3
|
| 218 |
+
}
|
| 219 |
+
},
|
| 220 |
+
"TrainerControl": {
|
| 221 |
+
"args": {
|
| 222 |
+
"should_epoch_stop": false,
|
| 223 |
+
"should_evaluate": false,
|
| 224 |
+
"should_log": false,
|
| 225 |
+
"should_save": true,
|
| 226 |
+
"should_training_stop": true
|
| 227 |
+
},
|
| 228 |
+
"attributes": {}
|
| 229 |
+
}
|
| 230 |
+
},
|
| 231 |
+
"total_flos": 0.0,
|
| 232 |
+
"train_batch_size": 8,
|
| 233 |
+
"trial_name": null,
|
| 234 |
+
"trial_params": null
|
| 235 |
+
}
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:349135d230d1ed107ab68d23c12e1ec29e33ead7aa6bb5d9825a51628b6a7642
|
| 3 |
+
size 5560
|
viclickbait2025-phobert-s42-20260811T112433Z-fe163e57/trainer/checkpoint-450/vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
viclickbait2025-phobert-s43-20260811T115543Z-6785dbb2/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d808facfe6618ae6ca7cb741532313a93d3a201d432e821bfb4359cea7178a8
|
| 3 |
+
size 540023320
|