medium model training 1
#3
by
camerxnlol
- opened
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/config.json +32 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/generation_config.json +6 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/optimizer.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/pytorch_model.bin +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/rng_state.pth +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/scheduler.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/trainer_state.json +733 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-1000/training_args.bin +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/config.json +32 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/generation_config.json +6 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/optimizer.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/pytorch_model.bin +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/rng_state.pth +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/scheduler.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/trainer_state.json +1433 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-2000/training_args.bin +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/config.json +32 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/generation_config.json +6 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/optimizer.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/pytorch_model.bin +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/rng_state.pth +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/scheduler.pt +3 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/trainer_state.json +2133 -0
- amt_PKMN_Harmonizer_Medium/checkpoint-3000/training_args.bin +3 -0
amt_PKMN_Harmonizer_Medium/checkpoint-1000/config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "stanford-crfm/music-medium-800k",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 55025,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 55025,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"layer_norm_epsilon": 1e-05,
|
| 13 |
+
"model_type": "gpt2",
|
| 14 |
+
"n_embd": 1024,
|
| 15 |
+
"n_head": 16,
|
| 16 |
+
"n_inner": null,
|
| 17 |
+
"n_layer": 24,
|
| 18 |
+
"n_positions": 1024,
|
| 19 |
+
"reorder_and_upcast_attn": false,
|
| 20 |
+
"resid_pdrop": 0.1,
|
| 21 |
+
"scale_attn_by_inverse_layer_idx": true,
|
| 22 |
+
"scale_attn_weights": true,
|
| 23 |
+
"summary_activation": null,
|
| 24 |
+
"summary_first_dropout": 0.1,
|
| 25 |
+
"summary_proj_to_labels": true,
|
| 26 |
+
"summary_type": "cls_index",
|
| 27 |
+
"summary_use_proj": true,
|
| 28 |
+
"torch_dtype": "float32",
|
| 29 |
+
"transformers_version": "4.49.0",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 55028
|
| 32 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 55025,
|
| 4 |
+
"eos_token_id": 55025,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf5ec13ce75ebfda557b874e2c46373b939d881c680e25fe13435f173c8ab333
|
| 3 |
+
size 2877913611
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8628b548555880f9a7bceba9cf5f1e5b753f7672984777d9cea6ee480c1bccda
|
| 3 |
+
size 1438930739
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:570b5def49842a667c05d912692aca3d2b49a73141b2daf9bb26577ecbd71c05
|
| 3 |
+
size 14645
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8223e0cce6fdd5319f657d06b97e0a68e98abbc917f32fdf3341592c98fd883d
|
| 3 |
+
size 1401
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,733 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.1154446177847115,
|
| 5 |
+
"eval_steps": 2000,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.031201248049921998,
|
| 13 |
+
"grad_norm": 59933.58984375,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 13.4575,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.062402496099843996,
|
| 20 |
+
"grad_norm": 79273.921875,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 13.0486,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.093603744149766,
|
| 27 |
+
"grad_norm": 86346.1953125,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 11.9289,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12480499219968799,
|
| 34 |
+
"grad_norm": 68253.6484375,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.78,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15600624024961,
|
| 41 |
+
"grad_norm": 48220.171875,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.4398,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.187207488299532,
|
| 48 |
+
"grad_norm": 31435.85546875,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 9.2586,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21840873634945399,
|
| 55 |
+
"grad_norm": 11692.8125,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 8.4688,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24960998439937598,
|
| 62 |
+
"grad_norm": 33474.359375,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 7.8024,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.28081123244929795,
|
| 69 |
+
"grad_norm": 11187.919921875,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 7.5132,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.31201248049922,
|
| 76 |
+
"grad_norm": 6494.74609375,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 7.0462,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.34321372854914195,
|
| 83 |
+
"grad_norm": 4932.1884765625,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 6.6515,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.374414976599064,
|
| 90 |
+
"grad_norm": 14954.1064453125,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 6.352,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.40561622464898595,
|
| 97 |
+
"grad_norm": 8677.0498046875,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 5.9368,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.43681747269890797,
|
| 104 |
+
"grad_norm": 4330.2470703125,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 5.6135,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.46801872074882994,
|
| 111 |
+
"grad_norm": 4594.47216796875,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 5.3826,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.49921996879875197,
|
| 118 |
+
"grad_norm": 4907.52880859375,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 5.1467,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5304212168486739,
|
| 125 |
+
"grad_norm": 2622.095458984375,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 5.0465,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5616224648985959,
|
| 132 |
+
"grad_norm": 2206.215087890625,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 4.868,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.592823712948518,
|
| 139 |
+
"grad_norm": 1866.2569580078125,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 4.7356,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.62402496099844,
|
| 146 |
+
"grad_norm": 1359.1365966796875,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 4.4869,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6552262090483619,
|
| 153 |
+
"grad_norm": 929.0238647460938,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 4.2993,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6864274570982839,
|
| 160 |
+
"grad_norm": 814.4940795898438,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 4.0853,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.717628705148206,
|
| 167 |
+
"grad_norm": 561.3009033203125,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 3.9475,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.748829953198128,
|
| 174 |
+
"grad_norm": 595.680908203125,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 3.8025,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7800312012480499,
|
| 181 |
+
"grad_norm": 713.8980102539062,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 3.6036,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8112324492979719,
|
| 188 |
+
"grad_norm": 373.4093017578125,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 3.4541,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8424336973478939,
|
| 195 |
+
"grad_norm": 7371.640625,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 3.3097,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8736349453978159,
|
| 202 |
+
"grad_norm": 380.82952880859375,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 3.2375,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9048361934477379,
|
| 209 |
+
"grad_norm": 418.1549377441406,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 3.1552,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9360374414976599,
|
| 216 |
+
"grad_norm": 884.2369384765625,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 3.0261,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9672386895475819,
|
| 223 |
+
"grad_norm": 425.3170166015625,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 2.9638,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9984399375975039,
|
| 230 |
+
"grad_norm": 346.85650634765625,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 2.939,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0280811232449298,
|
| 237 |
+
"grad_norm": 607.9754028320312,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 2.8921,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0592823712948518,
|
| 244 |
+
"grad_norm": 73.27684020996094,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 2.802,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0904836193447738,
|
| 251 |
+
"grad_norm": 77.26670837402344,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 2.6704,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1216848673946958,
|
| 258 |
+
"grad_norm": 46.717529296875,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 2.5633,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1528861154446177,
|
| 265 |
+
"grad_norm": 73.30224609375,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 2.4874,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1840873634945397,
|
| 272 |
+
"grad_norm": 82.5370101928711,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 2.4168,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2152886115444619,
|
| 279 |
+
"grad_norm": 56.439327239990234,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 2.2979,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2464898595943839,
|
| 286 |
+
"grad_norm": 34.27633285522461,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 2.2299,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2776911076443058,
|
| 293 |
+
"grad_norm": 24.567216873168945,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 2.1672,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3088923556942278,
|
| 300 |
+
"grad_norm": 28.32038688659668,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 2.1228,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3400936037441498,
|
| 307 |
+
"grad_norm": 47.05001449584961,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 2.0671,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3712948517940717,
|
| 314 |
+
"grad_norm": 59.189292907714844,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 1.9665,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4024960998439937,
|
| 321 |
+
"grad_norm": 62.23638916015625,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 1.9069,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4336973478939157,
|
| 328 |
+
"grad_norm": 20.689085006713867,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 1.8617,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4648985959438376,
|
| 335 |
+
"grad_norm": 21.1517333984375,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 1.8312,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4960998439937598,
|
| 342 |
+
"grad_norm": 11.305088996887207,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 1.7294,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5273010920436816,
|
| 349 |
+
"grad_norm": 10.452417373657227,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 1.641,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5585023400936038,
|
| 356 |
+
"grad_norm": 12.5308198928833,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 1.6129,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5897035881435257,
|
| 363 |
+
"grad_norm": 20.75917625427246,
|
| 364 |
+
"learning_rate": 9.999605221019082e-06,
|
| 365 |
+
"loss": 1.5093,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6209048361934477,
|
| 370 |
+
"grad_norm": 76.15613555908203,
|
| 371 |
+
"learning_rate": 9.9984209464165e-06,
|
| 372 |
+
"loss": 1.5462,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6521060842433697,
|
| 377 |
+
"grad_norm": 27.704729080200195,
|
| 378 |
+
"learning_rate": 9.996447363202947e-06,
|
| 379 |
+
"loss": 1.4907,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6833073322932917,
|
| 384 |
+
"grad_norm": 11.623372077941895,
|
| 385 |
+
"learning_rate": 9.99368478303009e-06,
|
| 386 |
+
"loss": 1.4383,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7145085803432139,
|
| 391 |
+
"grad_norm": 9.811054229736328,
|
| 392 |
+
"learning_rate": 9.990133642141359e-06,
|
| 393 |
+
"loss": 1.4416,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7457098283931356,
|
| 398 |
+
"grad_norm": 12.728073120117188,
|
| 399 |
+
"learning_rate": 9.98579450130307e-06,
|
| 400 |
+
"loss": 1.3726,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7769110764430578,
|
| 405 |
+
"grad_norm": 9.254298210144043,
|
| 406 |
+
"learning_rate": 9.980668045715864e-06,
|
| 407 |
+
"loss": 1.345,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8081123244929798,
|
| 412 |
+
"grad_norm": 19.775976181030273,
|
| 413 |
+
"learning_rate": 9.974755084906503e-06,
|
| 414 |
+
"loss": 1.3242,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8393135725429017,
|
| 419 |
+
"grad_norm": 6.770556926727295,
|
| 420 |
+
"learning_rate": 9.968056552600043e-06,
|
| 421 |
+
"loss": 1.292,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8705148205928237,
|
| 426 |
+
"grad_norm": 7.248960494995117,
|
| 427 |
+
"learning_rate": 9.960573506572391e-06,
|
| 428 |
+
"loss": 1.2693,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9017160686427457,
|
| 433 |
+
"grad_norm": 4.084126949310303,
|
| 434 |
+
"learning_rate": 9.952307128483257e-06,
|
| 435 |
+
"loss": 1.2154,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9329173166926679,
|
| 440 |
+
"grad_norm": 4.399320602416992,
|
| 441 |
+
"learning_rate": 9.94325872368957e-06,
|
| 442 |
+
"loss": 1.1869,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9641185647425896,
|
| 447 |
+
"grad_norm": 5.116558074951172,
|
| 448 |
+
"learning_rate": 9.93342972103934e-06,
|
| 449 |
+
"loss": 1.2168,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9953198127925118,
|
| 454 |
+
"grad_norm": 5.786620616912842,
|
| 455 |
+
"learning_rate": 9.922821672646028e-06,
|
| 456 |
+
"loss": 1.1734,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.0249609984399375,
|
| 461 |
+
"grad_norm": 15.972105979919434,
|
| 462 |
+
"learning_rate": 9.911436253643445e-06,
|
| 463 |
+
"loss": 1.1523,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0561622464898597,
|
| 468 |
+
"grad_norm": 4.64152193069458,
|
| 469 |
+
"learning_rate": 9.899275261921236e-06,
|
| 470 |
+
"loss": 1.1247,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0873634945397814,
|
| 475 |
+
"grad_norm": 7.009069442749023,
|
| 476 |
+
"learning_rate": 9.886340617840968e-06,
|
| 477 |
+
"loss": 1.0985,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.1185647425897036,
|
| 482 |
+
"grad_norm": 4.230483531951904,
|
| 483 |
+
"learning_rate": 9.872634363932887e-06,
|
| 484 |
+
"loss": 1.1014,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.149765990639626,
|
| 489 |
+
"grad_norm": 3.0746848583221436,
|
| 490 |
+
"learning_rate": 9.85815866457337e-06,
|
| 491 |
+
"loss": 1.0999,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1809672386895476,
|
| 496 |
+
"grad_norm": 3.9389874935150146,
|
| 497 |
+
"learning_rate": 9.842915805643156e-06,
|
| 498 |
+
"loss": 1.0328,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2121684867394698,
|
| 503 |
+
"grad_norm": 3.179098606109619,
|
| 504 |
+
"learning_rate": 9.82690819416637e-06,
|
| 505 |
+
"loss": 1.0448,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2433697347893915,
|
| 510 |
+
"grad_norm": 3.1507132053375244,
|
| 511 |
+
"learning_rate": 9.81013835793043e-06,
|
| 512 |
+
"loss": 1.0195,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.2745709828393137,
|
| 517 |
+
"grad_norm": 3.040616273880005,
|
| 518 |
+
"learning_rate": 9.79260894508688e-06,
|
| 519 |
+
"loss": 0.9717,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.3057722308892354,
|
| 524 |
+
"grad_norm": 3.114654779434204,
|
| 525 |
+
"learning_rate": 9.774322723733216e-06,
|
| 526 |
+
"loss": 0.9784,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.3369734789391576,
|
| 531 |
+
"grad_norm": 4.01024866104126,
|
| 532 |
+
"learning_rate": 9.755282581475769e-06,
|
| 533 |
+
"loss": 0.9872,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3681747269890794,
|
| 538 |
+
"grad_norm": 3.4513773918151855,
|
| 539 |
+
"learning_rate": 9.735491524973723e-06,
|
| 540 |
+
"loss": 0.9588,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3993759750390016,
|
| 545 |
+
"grad_norm": 3.086181879043579,
|
| 546 |
+
"learning_rate": 9.714952679464324e-06,
|
| 547 |
+
"loss": 0.9287,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4305772230889238,
|
| 552 |
+
"grad_norm": 2.971944570541382,
|
| 553 |
+
"learning_rate": 9.693669288269371e-06,
|
| 554 |
+
"loss": 0.9471,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.4617784711388455,
|
| 559 |
+
"grad_norm": 3.5111756324768066,
|
| 560 |
+
"learning_rate": 9.671644712283061e-06,
|
| 561 |
+
"loss": 0.9091,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4929797191887677,
|
| 566 |
+
"grad_norm": 3.4968149662017822,
|
| 567 |
+
"learning_rate": 9.648882429441258e-06,
|
| 568 |
+
"loss": 0.9133,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.5241809672386895,
|
| 573 |
+
"grad_norm": 2.7986488342285156,
|
| 574 |
+
"learning_rate": 9.62538603417229e-06,
|
| 575 |
+
"loss": 0.8803,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.5553822152886116,
|
| 580 |
+
"grad_norm": 2.635054349899292,
|
| 581 |
+
"learning_rate": 9.601159236829353e-06,
|
| 582 |
+
"loss": 0.904,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5865834633385334,
|
| 587 |
+
"grad_norm": 2.8344554901123047,
|
| 588 |
+
"learning_rate": 9.576205863104588e-06,
|
| 589 |
+
"loss": 0.8549,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.6177847113884556,
|
| 594 |
+
"grad_norm": 29.804149627685547,
|
| 595 |
+
"learning_rate": 9.550529853424979e-06,
|
| 596 |
+
"loss": 0.8314,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.6489859594383773,
|
| 601 |
+
"grad_norm": 4.379691123962402,
|
| 602 |
+
"learning_rate": 9.524135262330098e-06,
|
| 603 |
+
"loss": 0.8286,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6801872074882995,
|
| 608 |
+
"grad_norm": 3.7554783821105957,
|
| 609 |
+
"learning_rate": 9.497026257831856e-06,
|
| 610 |
+
"loss": 0.8682,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.7113884555382217,
|
| 615 |
+
"grad_norm": 2.8485891819000244,
|
| 616 |
+
"learning_rate": 9.46920712075632e-06,
|
| 617 |
+
"loss": 0.8415,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.7425897035881435,
|
| 622 |
+
"grad_norm": 2.6579718589782715,
|
| 623 |
+
"learning_rate": 9.440682244067724e-06,
|
| 624 |
+
"loss": 0.8467,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7737909516380657,
|
| 629 |
+
"grad_norm": 3.369091272354126,
|
| 630 |
+
"learning_rate": 9.411456132174768e-06,
|
| 631 |
+
"loss": 0.8357,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8049921996879874,
|
| 636 |
+
"grad_norm": 2.6015613079071045,
|
| 637 |
+
"learning_rate": 9.381533400219319e-06,
|
| 638 |
+
"loss": 0.8351,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8361934477379096,
|
| 643 |
+
"grad_norm": 3.204615354537964,
|
| 644 |
+
"learning_rate": 9.35091877334763e-06,
|
| 645 |
+
"loss": 0.8059,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.8673946957878313,
|
| 650 |
+
"grad_norm": 3.4444634914398193,
|
| 651 |
+
"learning_rate": 9.319617085964177e-06,
|
| 652 |
+
"loss": 0.7631,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8985959438377535,
|
| 657 |
+
"grad_norm": 2.5985376834869385,
|
| 658 |
+
"learning_rate": 9.287633280968263e-06,
|
| 659 |
+
"loss": 0.788,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.9297971918876753,
|
| 664 |
+
"grad_norm": 2.7666327953338623,
|
| 665 |
+
"learning_rate": 9.25497240897346e-06,
|
| 666 |
+
"loss": 0.7521,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.9609984399375975,
|
| 671 |
+
"grad_norm": 2.8274083137512207,
|
| 672 |
+
"learning_rate": 9.221639627510076e-06,
|
| 673 |
+
"loss": 0.7878,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9921996879875197,
|
| 678 |
+
"grad_norm": 2.522346019744873,
|
| 679 |
+
"learning_rate": 9.18764020021071e-06,
|
| 680 |
+
"loss": 0.7502,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0218408736349454,
|
| 685 |
+
"grad_norm": 2.3941545486450195,
|
| 686 |
+
"learning_rate": 9.152979495979064e-06,
|
| 687 |
+
"loss": 0.7451,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.0530421216848675,
|
| 692 |
+
"grad_norm": 3.3544349670410156,
|
| 693 |
+
"learning_rate": 9.117662988142138e-06,
|
| 694 |
+
"loss": 0.7193,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0842433697347893,
|
| 699 |
+
"grad_norm": 3.1165342330932617,
|
| 700 |
+
"learning_rate": 9.08169625358592e-06,
|
| 701 |
+
"loss": 0.7502,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.1154446177847115,
|
| 706 |
+
"grad_norm": 2.7650063037872314,
|
| 707 |
+
"learning_rate": 9.045084971874738e-06,
|
| 708 |
+
"loss": 0.7118,
|
| 709 |
+
"step": 1000
|
| 710 |
+
}
|
| 711 |
+
],
|
| 712 |
+
"logging_steps": 10,
|
| 713 |
+
"max_steps": 3000,
|
| 714 |
+
"num_input_tokens_seen": 0,
|
| 715 |
+
"num_train_epochs": 10,
|
| 716 |
+
"save_steps": 1000,
|
| 717 |
+
"stateful_callbacks": {
|
| 718 |
+
"TrainerControl": {
|
| 719 |
+
"args": {
|
| 720 |
+
"should_epoch_stop": false,
|
| 721 |
+
"should_evaluate": false,
|
| 722 |
+
"should_log": false,
|
| 723 |
+
"should_save": true,
|
| 724 |
+
"should_training_stop": false
|
| 725 |
+
},
|
| 726 |
+
"attributes": {}
|
| 727 |
+
}
|
| 728 |
+
},
|
| 729 |
+
"total_flos": 2.9673844591558656e+16,
|
| 730 |
+
"train_batch_size": 4,
|
| 731 |
+
"trial_name": null,
|
| 732 |
+
"trial_params": null
|
| 733 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:620747907b2130f2f54d285e74e3eb145407d8bba5652ffcf90771ba77206c41
|
| 3 |
+
size 5713
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "stanford-crfm/music-medium-800k",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 55025,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 55025,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"layer_norm_epsilon": 1e-05,
|
| 13 |
+
"model_type": "gpt2",
|
| 14 |
+
"n_embd": 1024,
|
| 15 |
+
"n_head": 16,
|
| 16 |
+
"n_inner": null,
|
| 17 |
+
"n_layer": 24,
|
| 18 |
+
"n_positions": 1024,
|
| 19 |
+
"reorder_and_upcast_attn": false,
|
| 20 |
+
"resid_pdrop": 0.1,
|
| 21 |
+
"scale_attn_by_inverse_layer_idx": true,
|
| 22 |
+
"scale_attn_weights": true,
|
| 23 |
+
"summary_activation": null,
|
| 24 |
+
"summary_first_dropout": 0.1,
|
| 25 |
+
"summary_proj_to_labels": true,
|
| 26 |
+
"summary_type": "cls_index",
|
| 27 |
+
"summary_use_proj": true,
|
| 28 |
+
"torch_dtype": "float32",
|
| 29 |
+
"transformers_version": "4.49.0",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 55028
|
| 32 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 55025,
|
| 4 |
+
"eos_token_id": 55025,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df1a03c79c244ca8a42c318325f48b01bda54917f94180a311a6e81ab9c0ec33
|
| 3 |
+
size 2877913611
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c1969531e7d01d43850d09b63f3d404dfe9d7a2383bcd2b91437a80b008e207
|
| 3 |
+
size 1438930739
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:601aeda531a6508249ab88e8713ee7d4893c5321c60c83bf2cb431577072c4a3
|
| 3 |
+
size 14645
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95e130ee4c62d3d3f0f706217b1ac4fabdde76437e79cc2a35ec47ea7de804c3
|
| 3 |
+
size 1401
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/trainer_state.json
ADDED
|
@@ -0,0 +1,1433 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 6.230889235569423,
|
| 5 |
+
"eval_steps": 2000,
|
| 6 |
+
"global_step": 2000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.031201248049921998,
|
| 13 |
+
"grad_norm": 59933.58984375,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 13.4575,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.062402496099843996,
|
| 20 |
+
"grad_norm": 79273.921875,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 13.0486,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.093603744149766,
|
| 27 |
+
"grad_norm": 86346.1953125,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 11.9289,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12480499219968799,
|
| 34 |
+
"grad_norm": 68253.6484375,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.78,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15600624024961,
|
| 41 |
+
"grad_norm": 48220.171875,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.4398,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.187207488299532,
|
| 48 |
+
"grad_norm": 31435.85546875,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 9.2586,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21840873634945399,
|
| 55 |
+
"grad_norm": 11692.8125,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 8.4688,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24960998439937598,
|
| 62 |
+
"grad_norm": 33474.359375,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 7.8024,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.28081123244929795,
|
| 69 |
+
"grad_norm": 11187.919921875,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 7.5132,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.31201248049922,
|
| 76 |
+
"grad_norm": 6494.74609375,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 7.0462,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.34321372854914195,
|
| 83 |
+
"grad_norm": 4932.1884765625,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 6.6515,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.374414976599064,
|
| 90 |
+
"grad_norm": 14954.1064453125,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 6.352,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.40561622464898595,
|
| 97 |
+
"grad_norm": 8677.0498046875,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 5.9368,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.43681747269890797,
|
| 104 |
+
"grad_norm": 4330.2470703125,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 5.6135,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.46801872074882994,
|
| 111 |
+
"grad_norm": 4594.47216796875,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 5.3826,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.49921996879875197,
|
| 118 |
+
"grad_norm": 4907.52880859375,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 5.1467,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5304212168486739,
|
| 125 |
+
"grad_norm": 2622.095458984375,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 5.0465,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5616224648985959,
|
| 132 |
+
"grad_norm": 2206.215087890625,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 4.868,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.592823712948518,
|
| 139 |
+
"grad_norm": 1866.2569580078125,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 4.7356,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.62402496099844,
|
| 146 |
+
"grad_norm": 1359.1365966796875,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 4.4869,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6552262090483619,
|
| 153 |
+
"grad_norm": 929.0238647460938,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 4.2993,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6864274570982839,
|
| 160 |
+
"grad_norm": 814.4940795898438,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 4.0853,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.717628705148206,
|
| 167 |
+
"grad_norm": 561.3009033203125,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 3.9475,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.748829953198128,
|
| 174 |
+
"grad_norm": 595.680908203125,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 3.8025,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7800312012480499,
|
| 181 |
+
"grad_norm": 713.8980102539062,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 3.6036,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8112324492979719,
|
| 188 |
+
"grad_norm": 373.4093017578125,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 3.4541,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8424336973478939,
|
| 195 |
+
"grad_norm": 7371.640625,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 3.3097,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8736349453978159,
|
| 202 |
+
"grad_norm": 380.82952880859375,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 3.2375,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9048361934477379,
|
| 209 |
+
"grad_norm": 418.1549377441406,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 3.1552,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9360374414976599,
|
| 216 |
+
"grad_norm": 884.2369384765625,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 3.0261,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9672386895475819,
|
| 223 |
+
"grad_norm": 425.3170166015625,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 2.9638,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9984399375975039,
|
| 230 |
+
"grad_norm": 346.85650634765625,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 2.939,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0280811232449298,
|
| 237 |
+
"grad_norm": 607.9754028320312,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 2.8921,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0592823712948518,
|
| 244 |
+
"grad_norm": 73.27684020996094,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 2.802,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0904836193447738,
|
| 251 |
+
"grad_norm": 77.26670837402344,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 2.6704,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1216848673946958,
|
| 258 |
+
"grad_norm": 46.717529296875,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 2.5633,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1528861154446177,
|
| 265 |
+
"grad_norm": 73.30224609375,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 2.4874,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1840873634945397,
|
| 272 |
+
"grad_norm": 82.5370101928711,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 2.4168,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2152886115444619,
|
| 279 |
+
"grad_norm": 56.439327239990234,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 2.2979,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2464898595943839,
|
| 286 |
+
"grad_norm": 34.27633285522461,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 2.2299,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2776911076443058,
|
| 293 |
+
"grad_norm": 24.567216873168945,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 2.1672,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3088923556942278,
|
| 300 |
+
"grad_norm": 28.32038688659668,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 2.1228,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3400936037441498,
|
| 307 |
+
"grad_norm": 47.05001449584961,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 2.0671,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3712948517940717,
|
| 314 |
+
"grad_norm": 59.189292907714844,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 1.9665,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4024960998439937,
|
| 321 |
+
"grad_norm": 62.23638916015625,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 1.9069,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4336973478939157,
|
| 328 |
+
"grad_norm": 20.689085006713867,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 1.8617,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4648985959438376,
|
| 335 |
+
"grad_norm": 21.1517333984375,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 1.8312,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4960998439937598,
|
| 342 |
+
"grad_norm": 11.305088996887207,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 1.7294,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5273010920436816,
|
| 349 |
+
"grad_norm": 10.452417373657227,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 1.641,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5585023400936038,
|
| 356 |
+
"grad_norm": 12.5308198928833,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 1.6129,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5897035881435257,
|
| 363 |
+
"grad_norm": 20.75917625427246,
|
| 364 |
+
"learning_rate": 9.999605221019082e-06,
|
| 365 |
+
"loss": 1.5093,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6209048361934477,
|
| 370 |
+
"grad_norm": 76.15613555908203,
|
| 371 |
+
"learning_rate": 9.9984209464165e-06,
|
| 372 |
+
"loss": 1.5462,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6521060842433697,
|
| 377 |
+
"grad_norm": 27.704729080200195,
|
| 378 |
+
"learning_rate": 9.996447363202947e-06,
|
| 379 |
+
"loss": 1.4907,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6833073322932917,
|
| 384 |
+
"grad_norm": 11.623372077941895,
|
| 385 |
+
"learning_rate": 9.99368478303009e-06,
|
| 386 |
+
"loss": 1.4383,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7145085803432139,
|
| 391 |
+
"grad_norm": 9.811054229736328,
|
| 392 |
+
"learning_rate": 9.990133642141359e-06,
|
| 393 |
+
"loss": 1.4416,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7457098283931356,
|
| 398 |
+
"grad_norm": 12.728073120117188,
|
| 399 |
+
"learning_rate": 9.98579450130307e-06,
|
| 400 |
+
"loss": 1.3726,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7769110764430578,
|
| 405 |
+
"grad_norm": 9.254298210144043,
|
| 406 |
+
"learning_rate": 9.980668045715864e-06,
|
| 407 |
+
"loss": 1.345,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8081123244929798,
|
| 412 |
+
"grad_norm": 19.775976181030273,
|
| 413 |
+
"learning_rate": 9.974755084906503e-06,
|
| 414 |
+
"loss": 1.3242,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8393135725429017,
|
| 419 |
+
"grad_norm": 6.770556926727295,
|
| 420 |
+
"learning_rate": 9.968056552600043e-06,
|
| 421 |
+
"loss": 1.292,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8705148205928237,
|
| 426 |
+
"grad_norm": 7.248960494995117,
|
| 427 |
+
"learning_rate": 9.960573506572391e-06,
|
| 428 |
+
"loss": 1.2693,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9017160686427457,
|
| 433 |
+
"grad_norm": 4.084126949310303,
|
| 434 |
+
"learning_rate": 9.952307128483257e-06,
|
| 435 |
+
"loss": 1.2154,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9329173166926679,
|
| 440 |
+
"grad_norm": 4.399320602416992,
|
| 441 |
+
"learning_rate": 9.94325872368957e-06,
|
| 442 |
+
"loss": 1.1869,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9641185647425896,
|
| 447 |
+
"grad_norm": 5.116558074951172,
|
| 448 |
+
"learning_rate": 9.93342972103934e-06,
|
| 449 |
+
"loss": 1.2168,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9953198127925118,
|
| 454 |
+
"grad_norm": 5.786620616912842,
|
| 455 |
+
"learning_rate": 9.922821672646028e-06,
|
| 456 |
+
"loss": 1.1734,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.0249609984399375,
|
| 461 |
+
"grad_norm": 15.972105979919434,
|
| 462 |
+
"learning_rate": 9.911436253643445e-06,
|
| 463 |
+
"loss": 1.1523,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0561622464898597,
|
| 468 |
+
"grad_norm": 4.64152193069458,
|
| 469 |
+
"learning_rate": 9.899275261921236e-06,
|
| 470 |
+
"loss": 1.1247,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0873634945397814,
|
| 475 |
+
"grad_norm": 7.009069442749023,
|
| 476 |
+
"learning_rate": 9.886340617840968e-06,
|
| 477 |
+
"loss": 1.0985,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.1185647425897036,
|
| 482 |
+
"grad_norm": 4.230483531951904,
|
| 483 |
+
"learning_rate": 9.872634363932887e-06,
|
| 484 |
+
"loss": 1.1014,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.149765990639626,
|
| 489 |
+
"grad_norm": 3.0746848583221436,
|
| 490 |
+
"learning_rate": 9.85815866457337e-06,
|
| 491 |
+
"loss": 1.0999,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1809672386895476,
|
| 496 |
+
"grad_norm": 3.9389874935150146,
|
| 497 |
+
"learning_rate": 9.842915805643156e-06,
|
| 498 |
+
"loss": 1.0328,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2121684867394698,
|
| 503 |
+
"grad_norm": 3.179098606109619,
|
| 504 |
+
"learning_rate": 9.82690819416637e-06,
|
| 505 |
+
"loss": 1.0448,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2433697347893915,
|
| 510 |
+
"grad_norm": 3.1507132053375244,
|
| 511 |
+
"learning_rate": 9.81013835793043e-06,
|
| 512 |
+
"loss": 1.0195,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.2745709828393137,
|
| 517 |
+
"grad_norm": 3.040616273880005,
|
| 518 |
+
"learning_rate": 9.79260894508688e-06,
|
| 519 |
+
"loss": 0.9717,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.3057722308892354,
|
| 524 |
+
"grad_norm": 3.114654779434204,
|
| 525 |
+
"learning_rate": 9.774322723733216e-06,
|
| 526 |
+
"loss": 0.9784,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.3369734789391576,
|
| 531 |
+
"grad_norm": 4.01024866104126,
|
| 532 |
+
"learning_rate": 9.755282581475769e-06,
|
| 533 |
+
"loss": 0.9872,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3681747269890794,
|
| 538 |
+
"grad_norm": 3.4513773918151855,
|
| 539 |
+
"learning_rate": 9.735491524973723e-06,
|
| 540 |
+
"loss": 0.9588,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3993759750390016,
|
| 545 |
+
"grad_norm": 3.086181879043579,
|
| 546 |
+
"learning_rate": 9.714952679464324e-06,
|
| 547 |
+
"loss": 0.9287,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4305772230889238,
|
| 552 |
+
"grad_norm": 2.971944570541382,
|
| 553 |
+
"learning_rate": 9.693669288269371e-06,
|
| 554 |
+
"loss": 0.9471,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.4617784711388455,
|
| 559 |
+
"grad_norm": 3.5111756324768066,
|
| 560 |
+
"learning_rate": 9.671644712283061e-06,
|
| 561 |
+
"loss": 0.9091,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4929797191887677,
|
| 566 |
+
"grad_norm": 3.4968149662017822,
|
| 567 |
+
"learning_rate": 9.648882429441258e-06,
|
| 568 |
+
"loss": 0.9133,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.5241809672386895,
|
| 573 |
+
"grad_norm": 2.7986488342285156,
|
| 574 |
+
"learning_rate": 9.62538603417229e-06,
|
| 575 |
+
"loss": 0.8803,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.5553822152886116,
|
| 580 |
+
"grad_norm": 2.635054349899292,
|
| 581 |
+
"learning_rate": 9.601159236829353e-06,
|
| 582 |
+
"loss": 0.904,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5865834633385334,
|
| 587 |
+
"grad_norm": 2.8344554901123047,
|
| 588 |
+
"learning_rate": 9.576205863104588e-06,
|
| 589 |
+
"loss": 0.8549,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.6177847113884556,
|
| 594 |
+
"grad_norm": 29.804149627685547,
|
| 595 |
+
"learning_rate": 9.550529853424979e-06,
|
| 596 |
+
"loss": 0.8314,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.6489859594383773,
|
| 601 |
+
"grad_norm": 4.379691123962402,
|
| 602 |
+
"learning_rate": 9.524135262330098e-06,
|
| 603 |
+
"loss": 0.8286,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6801872074882995,
|
| 608 |
+
"grad_norm": 3.7554783821105957,
|
| 609 |
+
"learning_rate": 9.497026257831856e-06,
|
| 610 |
+
"loss": 0.8682,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.7113884555382217,
|
| 615 |
+
"grad_norm": 2.8485891819000244,
|
| 616 |
+
"learning_rate": 9.46920712075632e-06,
|
| 617 |
+
"loss": 0.8415,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.7425897035881435,
|
| 622 |
+
"grad_norm": 2.6579718589782715,
|
| 623 |
+
"learning_rate": 9.440682244067724e-06,
|
| 624 |
+
"loss": 0.8467,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7737909516380657,
|
| 629 |
+
"grad_norm": 3.369091272354126,
|
| 630 |
+
"learning_rate": 9.411456132174768e-06,
|
| 631 |
+
"loss": 0.8357,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8049921996879874,
|
| 636 |
+
"grad_norm": 2.6015613079071045,
|
| 637 |
+
"learning_rate": 9.381533400219319e-06,
|
| 638 |
+
"loss": 0.8351,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8361934477379096,
|
| 643 |
+
"grad_norm": 3.204615354537964,
|
| 644 |
+
"learning_rate": 9.35091877334763e-06,
|
| 645 |
+
"loss": 0.8059,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.8673946957878313,
|
| 650 |
+
"grad_norm": 3.4444634914398193,
|
| 651 |
+
"learning_rate": 9.319617085964177e-06,
|
| 652 |
+
"loss": 0.7631,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8985959438377535,
|
| 657 |
+
"grad_norm": 2.5985376834869385,
|
| 658 |
+
"learning_rate": 9.287633280968263e-06,
|
| 659 |
+
"loss": 0.788,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.9297971918876753,
|
| 664 |
+
"grad_norm": 2.7666327953338623,
|
| 665 |
+
"learning_rate": 9.25497240897346e-06,
|
| 666 |
+
"loss": 0.7521,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.9609984399375975,
|
| 671 |
+
"grad_norm": 2.8274083137512207,
|
| 672 |
+
"learning_rate": 9.221639627510076e-06,
|
| 673 |
+
"loss": 0.7878,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9921996879875197,
|
| 678 |
+
"grad_norm": 2.522346019744873,
|
| 679 |
+
"learning_rate": 9.18764020021071e-06,
|
| 680 |
+
"loss": 0.7502,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0218408736349454,
|
| 685 |
+
"grad_norm": 2.3941545486450195,
|
| 686 |
+
"learning_rate": 9.152979495979064e-06,
|
| 687 |
+
"loss": 0.7451,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.0530421216848675,
|
| 692 |
+
"grad_norm": 3.3544349670410156,
|
| 693 |
+
"learning_rate": 9.117662988142138e-06,
|
| 694 |
+
"loss": 0.7193,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0842433697347893,
|
| 699 |
+
"grad_norm": 3.1165342330932617,
|
| 700 |
+
"learning_rate": 9.08169625358592e-06,
|
| 701 |
+
"loss": 0.7502,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.1154446177847115,
|
| 706 |
+
"grad_norm": 2.7650063037872314,
|
| 707 |
+
"learning_rate": 9.045084971874738e-06,
|
| 708 |
+
"loss": 0.7118,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.1466458658346332,
|
| 713 |
+
"grad_norm": 2.8753139972686768,
|
| 714 |
+
"learning_rate": 9.007834924354384e-06,
|
| 715 |
+
"loss": 0.7367,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.1778471138845554,
|
| 720 |
+
"grad_norm": 4.134742736816406,
|
| 721 |
+
"learning_rate": 8.969951993239177e-06,
|
| 722 |
+
"loss": 0.6477,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.209048361934477,
|
| 727 |
+
"grad_norm": 2.406000852584839,
|
| 728 |
+
"learning_rate": 8.931442160683094e-06,
|
| 729 |
+
"loss": 0.7034,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.2402496099843994,
|
| 734 |
+
"grad_norm": 2.5077903270721436,
|
| 735 |
+
"learning_rate": 8.892311507835118e-06,
|
| 736 |
+
"loss": 0.7009,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.2714508580343216,
|
| 741 |
+
"grad_norm": 2.7244954109191895,
|
| 742 |
+
"learning_rate": 8.852566213878947e-06,
|
| 743 |
+
"loss": 0.6858,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.3026521060842433,
|
| 748 |
+
"grad_norm": 2.5308499336242676,
|
| 749 |
+
"learning_rate": 8.81221255505724e-06,
|
| 750 |
+
"loss": 0.6717,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.3338533541341655,
|
| 755 |
+
"grad_norm": 2.368601083755493,
|
| 756 |
+
"learning_rate": 8.77125690368052e-06,
|
| 757 |
+
"loss": 0.6634,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.3650546021840873,
|
| 762 |
+
"grad_norm": 2.6542091369628906,
|
| 763 |
+
"learning_rate": 8.729705727120911e-06,
|
| 764 |
+
"loss": 0.6711,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.3962558502340094,
|
| 769 |
+
"grad_norm": 2.466643810272217,
|
| 770 |
+
"learning_rate": 8.68756558679087e-06,
|
| 771 |
+
"loss": 0.6418,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.427457098283931,
|
| 776 |
+
"grad_norm": 2.714127779006958,
|
| 777 |
+
"learning_rate": 8.644843137107058e-06,
|
| 778 |
+
"loss": 0.6524,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.4586583463338534,
|
| 783 |
+
"grad_norm": 2.508812427520752,
|
| 784 |
+
"learning_rate": 8.601545124439535e-06,
|
| 785 |
+
"loss": 0.6684,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.489859594383775,
|
| 790 |
+
"grad_norm": 2.2211365699768066,
|
| 791 |
+
"learning_rate": 8.557678386046427e-06,
|
| 792 |
+
"loss": 0.641,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.5210608424336973,
|
| 797 |
+
"grad_norm": 6.399374008178711,
|
| 798 |
+
"learning_rate": 8.513249848994248e-06,
|
| 799 |
+
"loss": 0.6831,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.5522620904836195,
|
| 804 |
+
"grad_norm": 2.5456409454345703,
|
| 805 |
+
"learning_rate": 8.468266529064025e-06,
|
| 806 |
+
"loss": 0.6274,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.5834633385335413,
|
| 811 |
+
"grad_norm": 2.545017719268799,
|
| 812 |
+
"learning_rate": 8.422735529643445e-06,
|
| 813 |
+
"loss": 0.6361,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.6146645865834635,
|
| 818 |
+
"grad_norm": 2.749608278274536,
|
| 819 |
+
"learning_rate": 8.376664040605122e-06,
|
| 820 |
+
"loss": 0.6233,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.645865834633385,
|
| 825 |
+
"grad_norm": 2.785017728805542,
|
| 826 |
+
"learning_rate": 8.33005933717126e-06,
|
| 827 |
+
"loss": 0.6125,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.6770670826833074,
|
| 832 |
+
"grad_norm": 2.522922992706299,
|
| 833 |
+
"learning_rate": 8.282928778764783e-06,
|
| 834 |
+
"loss": 0.5696,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.7082683307332296,
|
| 839 |
+
"grad_norm": 2.222412347793579,
|
| 840 |
+
"learning_rate": 8.235279807847223e-06,
|
| 841 |
+
"loss": 0.6293,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.7394695787831513,
|
| 846 |
+
"grad_norm": 2.597766399383545,
|
| 847 |
+
"learning_rate": 8.18711994874345e-06,
|
| 848 |
+
"loss": 0.6062,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.770670826833073,
|
| 853 |
+
"grad_norm": 2.3702147006988525,
|
| 854 |
+
"learning_rate": 8.138456806453503e-06,
|
| 855 |
+
"loss": 0.5789,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.8018720748829953,
|
| 860 |
+
"grad_norm": 4.38936185836792,
|
| 861 |
+
"learning_rate": 8.089298065451673e-06,
|
| 862 |
+
"loss": 0.6036,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.8330733229329175,
|
| 867 |
+
"grad_norm": 2.9343149662017822,
|
| 868 |
+
"learning_rate": 8.039651488473028e-06,
|
| 869 |
+
"loss": 0.5832,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.864274570982839,
|
| 874 |
+
"grad_norm": 2.2963383197784424,
|
| 875 |
+
"learning_rate": 7.989524915287595e-06,
|
| 876 |
+
"loss": 0.5975,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.8954758190327614,
|
| 881 |
+
"grad_norm": 7.433852672576904,
|
| 882 |
+
"learning_rate": 7.938926261462366e-06,
|
| 883 |
+
"loss": 0.594,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.926677067082683,
|
| 888 |
+
"grad_norm": 2.5729384422302246,
|
| 889 |
+
"learning_rate": 7.887863517111337e-06,
|
| 890 |
+
"loss": 0.5937,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.9578783151326054,
|
| 895 |
+
"grad_norm": 2.6089067459106445,
|
| 896 |
+
"learning_rate": 7.836344745633785e-06,
|
| 897 |
+
"loss": 0.6126,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.9890795631825275,
|
| 902 |
+
"grad_norm": 2.289501667022705,
|
| 903 |
+
"learning_rate": 7.78437808244094e-06,
|
| 904 |
+
"loss": 0.5892,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.018720748829953,
|
| 909 |
+
"grad_norm": 2.9094228744506836,
|
| 910 |
+
"learning_rate": 7.731971733671347e-06,
|
| 911 |
+
"loss": 0.5797,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.049921996879875,
|
| 916 |
+
"grad_norm": 2.0647175312042236,
|
| 917 |
+
"learning_rate": 7.679133974894984e-06,
|
| 918 |
+
"loss": 0.5628,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.081123244929797,
|
| 923 |
+
"grad_norm": 2.374997854232788,
|
| 924 |
+
"learning_rate": 7.6258731498064796e-06,
|
| 925 |
+
"loss": 0.5923,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.112324492979719,
|
| 930 |
+
"grad_norm": 2.2724082469940186,
|
| 931 |
+
"learning_rate": 7.572197668907533e-06,
|
| 932 |
+
"loss": 0.5574,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.1435257410296416,
|
| 937 |
+
"grad_norm": 2.1879308223724365,
|
| 938 |
+
"learning_rate": 7.518116008178805e-06,
|
| 939 |
+
"loss": 0.5568,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.174726989079563,
|
| 944 |
+
"grad_norm": 2.5050461292266846,
|
| 945 |
+
"learning_rate": 7.463636707741458e-06,
|
| 946 |
+
"loss": 0.526,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.205928237129485,
|
| 951 |
+
"grad_norm": 2.2555339336395264,
|
| 952 |
+
"learning_rate": 7.408768370508577e-06,
|
| 953 |
+
"loss": 0.5302,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.237129485179407,
|
| 958 |
+
"grad_norm": 2.5363669395446777,
|
| 959 |
+
"learning_rate": 7.353519660826665e-06,
|
| 960 |
+
"loss": 0.5485,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.268330733229329,
|
| 965 |
+
"grad_norm": 2.200976848602295,
|
| 966 |
+
"learning_rate": 7.297899303107441e-06,
|
| 967 |
+
"loss": 0.5301,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.299531981279252,
|
| 972 |
+
"grad_norm": 2.2207038402557373,
|
| 973 |
+
"learning_rate": 7.241916080450163e-06,
|
| 974 |
+
"loss": 0.51,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.330733229329173,
|
| 979 |
+
"grad_norm": 2.300055980682373,
|
| 980 |
+
"learning_rate": 7.185578833254665e-06,
|
| 981 |
+
"loss": 0.5216,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.361934477379095,
|
| 986 |
+
"grad_norm": 2.3568477630615234,
|
| 987 |
+
"learning_rate": 7.128896457825364e-06,
|
| 988 |
+
"loss": 0.545,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.393135725429017,
|
| 993 |
+
"grad_norm": 2.1130895614624023,
|
| 994 |
+
"learning_rate": 7.071877904966422e-06,
|
| 995 |
+
"loss": 0.5231,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.4243369734789395,
|
| 1000 |
+
"grad_norm": 2.3051438331604004,
|
| 1001 |
+
"learning_rate": 7.014532178568314e-06,
|
| 1002 |
+
"loss": 0.5072,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.455538221528861,
|
| 1007 |
+
"grad_norm": 2.8409862518310547,
|
| 1008 |
+
"learning_rate": 6.9568683341860135e-06,
|
| 1009 |
+
"loss": 0.4753,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.486739469578783,
|
| 1014 |
+
"grad_norm": 2.3482327461242676,
|
| 1015 |
+
"learning_rate": 6.898895477609007e-06,
|
| 1016 |
+
"loss": 0.5158,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.517940717628705,
|
| 1021 |
+
"grad_norm": 2.4122118949890137,
|
| 1022 |
+
"learning_rate": 6.840622763423391e-06,
|
| 1023 |
+
"loss": 0.518,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.549141965678627,
|
| 1028 |
+
"grad_norm": 2.0749752521514893,
|
| 1029 |
+
"learning_rate": 6.782059393566254e-06,
|
| 1030 |
+
"loss": 0.4895,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.580343213728549,
|
| 1035 |
+
"grad_norm": 2.186260461807251,
|
| 1036 |
+
"learning_rate": 6.723214615872585e-06,
|
| 1037 |
+
"loss": 0.4872,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.611544461778471,
|
| 1042 |
+
"grad_norm": 2.3305256366729736,
|
| 1043 |
+
"learning_rate": 6.664097722614934e-06,
|
| 1044 |
+
"loss": 0.4935,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.642745709828393,
|
| 1049 |
+
"grad_norm": 2.2758896350860596,
|
| 1050 |
+
"learning_rate": 6.604718049036047e-06,
|
| 1051 |
+
"loss": 0.4942,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.673946957878315,
|
| 1056 |
+
"grad_norm": 3.157252311706543,
|
| 1057 |
+
"learning_rate": 6.545084971874738e-06,
|
| 1058 |
+
"loss": 0.4939,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.7051482059282375,
|
| 1063 |
+
"grad_norm": 2.2830564975738525,
|
| 1064 |
+
"learning_rate": 6.485207907885175e-06,
|
| 1065 |
+
"loss": 0.5057,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.736349453978159,
|
| 1070 |
+
"grad_norm": 2.0698862075805664,
|
| 1071 |
+
"learning_rate": 6.425096312349881e-06,
|
| 1072 |
+
"loss": 0.4922,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.767550702028081,
|
| 1077 |
+
"grad_norm": 2.110299587249756,
|
| 1078 |
+
"learning_rate": 6.364759677586627e-06,
|
| 1079 |
+
"loss": 0.4821,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.798751950078003,
|
| 1084 |
+
"grad_norm": 2.074535369873047,
|
| 1085 |
+
"learning_rate": 6.304207531449486e-06,
|
| 1086 |
+
"loss": 0.4652,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.829953198127925,
|
| 1091 |
+
"grad_norm": 2.317800998687744,
|
| 1092 |
+
"learning_rate": 6.243449435824276e-06,
|
| 1093 |
+
"loss": 0.4867,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.8611544461778475,
|
| 1098 |
+
"grad_norm": 2.319498062133789,
|
| 1099 |
+
"learning_rate": 6.182494985118625e-06,
|
| 1100 |
+
"loss": 0.4316,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.892355694227769,
|
| 1105 |
+
"grad_norm": 2.2666006088256836,
|
| 1106 |
+
"learning_rate": 6.121353804746907e-06,
|
| 1107 |
+
"loss": 0.466,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.923556942277691,
|
| 1112 |
+
"grad_norm": 2.0361239910125732,
|
| 1113 |
+
"learning_rate": 6.060035549610275e-06,
|
| 1114 |
+
"loss": 0.4615,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.954758190327613,
|
| 1119 |
+
"grad_norm": 2.0396602153778076,
|
| 1120 |
+
"learning_rate": 5.9985499025720354e-06,
|
| 1121 |
+
"loss": 0.4672,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.985959438377535,
|
| 1126 |
+
"grad_norm": 1.9205994606018066,
|
| 1127 |
+
"learning_rate": 5.936906572928625e-06,
|
| 1128 |
+
"loss": 0.4197,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.015600624024961,
|
| 1133 |
+
"grad_norm": 2.4138858318328857,
|
| 1134 |
+
"learning_rate": 5.8751152948763815e-06,
|
| 1135 |
+
"loss": 0.4384,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.046801872074883,
|
| 1140 |
+
"grad_norm": 2.2438735961914062,
|
| 1141 |
+
"learning_rate": 5.813185825974419e-06,
|
| 1142 |
+
"loss": 0.4392,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.078003120124805,
|
| 1147 |
+
"grad_norm": 2.7369537353515625,
|
| 1148 |
+
"learning_rate": 5.751127945603786e-06,
|
| 1149 |
+
"loss": 0.4654,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.109204368174727,
|
| 1154 |
+
"grad_norm": 2.676726818084717,
|
| 1155 |
+
"learning_rate": 5.68895145342319e-06,
|
| 1156 |
+
"loss": 0.4572,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.140405616224649,
|
| 1161 |
+
"grad_norm": 2.2318201065063477,
|
| 1162 |
+
"learning_rate": 5.626666167821522e-06,
|
| 1163 |
+
"loss": 0.4801,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.171606864274571,
|
| 1168 |
+
"grad_norm": 2.003933906555176,
|
| 1169 |
+
"learning_rate": 5.5642819243674085e-06,
|
| 1170 |
+
"loss": 0.4197,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.202808112324493,
|
| 1175 |
+
"grad_norm": 1.9387847185134888,
|
| 1176 |
+
"learning_rate": 5.5018085742560745e-06,
|
| 1177 |
+
"loss": 0.4372,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.234009360374415,
|
| 1182 |
+
"grad_norm": 2.0050766468048096,
|
| 1183 |
+
"learning_rate": 5.439255982753717e-06,
|
| 1184 |
+
"loss": 0.4553,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.265210608424337,
|
| 1189 |
+
"grad_norm": 2.092911720275879,
|
| 1190 |
+
"learning_rate": 5.376634027639664e-06,
|
| 1191 |
+
"loss": 0.4383,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.296411856474259,
|
| 1196 |
+
"grad_norm": 1.9465577602386475,
|
| 1197 |
+
"learning_rate": 5.3139525976465675e-06,
|
| 1198 |
+
"loss": 0.4549,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.327613104524181,
|
| 1203 |
+
"grad_norm": 2.0983693599700928,
|
| 1204 |
+
"learning_rate": 5.251221590898848e-06,
|
| 1205 |
+
"loss": 0.4639,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.358814352574103,
|
| 1210 |
+
"grad_norm": 1.85776686668396,
|
| 1211 |
+
"learning_rate": 5.188450913349674e-06,
|
| 1212 |
+
"loss": 0.4129,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.390015600624025,
|
| 1217 |
+
"grad_norm": 1.9787360429763794,
|
| 1218 |
+
"learning_rate": 5.1256504772166885e-06,
|
| 1219 |
+
"loss": 0.4416,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.4212168486739465,
|
| 1224 |
+
"grad_norm": 2.25137996673584,
|
| 1225 |
+
"learning_rate": 5.062830199416764e-06,
|
| 1226 |
+
"loss": 0.4325,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.452418096723869,
|
| 1231 |
+
"grad_norm": 2.2280070781707764,
|
| 1232 |
+
"learning_rate": 5e-06,
|
| 1233 |
+
"loss": 0.4061,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.483619344773791,
|
| 1238 |
+
"grad_norm": 1.9798814058303833,
|
| 1239 |
+
"learning_rate": 4.937169800583237e-06,
|
| 1240 |
+
"loss": 0.4398,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.514820592823713,
|
| 1245 |
+
"grad_norm": 1.8920650482177734,
|
| 1246 |
+
"learning_rate": 4.874349522783313e-06,
|
| 1247 |
+
"loss": 0.4009,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.546021840873635,
|
| 1252 |
+
"grad_norm": 2.217628240585327,
|
| 1253 |
+
"learning_rate": 4.811549086650327e-06,
|
| 1254 |
+
"loss": 0.4084,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.577223088923557,
|
| 1259 |
+
"grad_norm": 2.321141481399536,
|
| 1260 |
+
"learning_rate": 4.748778409101153e-06,
|
| 1261 |
+
"loss": 0.4229,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.608424336973479,
|
| 1266 |
+
"grad_norm": 2.1298415660858154,
|
| 1267 |
+
"learning_rate": 4.686047402353433e-06,
|
| 1268 |
+
"loss": 0.3925,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.639625585023401,
|
| 1273 |
+
"grad_norm": 1.9593074321746826,
|
| 1274 |
+
"learning_rate": 4.6233659723603374e-06,
|
| 1275 |
+
"loss": 0.4292,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.670826833073323,
|
| 1280 |
+
"grad_norm": 2.3608038425445557,
|
| 1281 |
+
"learning_rate": 4.560744017246284e-06,
|
| 1282 |
+
"loss": 0.3864,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.702028081123245,
|
| 1287 |
+
"grad_norm": 1.8943296670913696,
|
| 1288 |
+
"learning_rate": 4.4981914257439254e-06,
|
| 1289 |
+
"loss": 0.4034,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.733229329173167,
|
| 1294 |
+
"grad_norm": 2.0361130237579346,
|
| 1295 |
+
"learning_rate": 4.4357180756325915e-06,
|
| 1296 |
+
"loss": 0.3829,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.764430577223089,
|
| 1301 |
+
"grad_norm": 1.977895975112915,
|
| 1302 |
+
"learning_rate": 4.373333832178478e-06,
|
| 1303 |
+
"loss": 0.4345,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.795631825273011,
|
| 1308 |
+
"grad_norm": 2.087878465652466,
|
| 1309 |
+
"learning_rate": 4.31104854657681e-06,
|
| 1310 |
+
"loss": 0.4115,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.826833073322933,
|
| 1315 |
+
"grad_norm": 2.121720314025879,
|
| 1316 |
+
"learning_rate": 4.248872054396215e-06,
|
| 1317 |
+
"loss": 0.3975,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.858034321372855,
|
| 1322 |
+
"grad_norm": 2.06178879737854,
|
| 1323 |
+
"learning_rate": 4.186814174025582e-06,
|
| 1324 |
+
"loss": 0.4194,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.889235569422777,
|
| 1329 |
+
"grad_norm": 1.8563209772109985,
|
| 1330 |
+
"learning_rate": 4.124884705123619e-06,
|
| 1331 |
+
"loss": 0.4003,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.920436817472699,
|
| 1336 |
+
"grad_norm": 2.121776580810547,
|
| 1337 |
+
"learning_rate": 4.063093427071376e-06,
|
| 1338 |
+
"loss": 0.4409,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.951638065522621,
|
| 1343 |
+
"grad_norm": 2.0366406440734863,
|
| 1344 |
+
"learning_rate": 4.001450097427965e-06,
|
| 1345 |
+
"loss": 0.3787,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.982839313572543,
|
| 1350 |
+
"grad_norm": 1.861203670501709,
|
| 1351 |
+
"learning_rate": 3.939964450389728e-06,
|
| 1352 |
+
"loss": 0.4429,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.0124804992199685,
|
| 1357 |
+
"grad_norm": 2.0258266925811768,
|
| 1358 |
+
"learning_rate": 3.8786461952530955e-06,
|
| 1359 |
+
"loss": 0.3855,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.043681747269891,
|
| 1364 |
+
"grad_norm": 1.9031906127929688,
|
| 1365 |
+
"learning_rate": 3.817505014881378e-06,
|
| 1366 |
+
"loss": 0.3763,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.074882995319813,
|
| 1371 |
+
"grad_norm": 2.047241449356079,
|
| 1372 |
+
"learning_rate": 3.756550564175727e-06,
|
| 1373 |
+
"loss": 0.3668,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.106084243369735,
|
| 1378 |
+
"grad_norm": 2.1982035636901855,
|
| 1379 |
+
"learning_rate": 3.695792468550517e-06,
|
| 1380 |
+
"loss": 0.3892,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.137285491419656,
|
| 1385 |
+
"grad_norm": 1.949191689491272,
|
| 1386 |
+
"learning_rate": 3.635240322413375e-06,
|
| 1387 |
+
"loss": 0.418,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.168486739469579,
|
| 1392 |
+
"grad_norm": 2.028028726577759,
|
| 1393 |
+
"learning_rate": 3.5749036876501196e-06,
|
| 1394 |
+
"loss": 0.3972,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.199687987519501,
|
| 1399 |
+
"grad_norm": 2.0211682319641113,
|
| 1400 |
+
"learning_rate": 3.5147920921148267e-06,
|
| 1401 |
+
"loss": 0.4032,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.230889235569423,
|
| 1406 |
+
"grad_norm": 2.1258091926574707,
|
| 1407 |
+
"learning_rate": 3.4549150281252635e-06,
|
| 1408 |
+
"loss": 0.3885,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
}
|
| 1411 |
+
],
|
| 1412 |
+
"logging_steps": 10,
|
| 1413 |
+
"max_steps": 3000,
|
| 1414 |
+
"num_input_tokens_seen": 0,
|
| 1415 |
+
"num_train_epochs": 10,
|
| 1416 |
+
"save_steps": 1000,
|
| 1417 |
+
"stateful_callbacks": {
|
| 1418 |
+
"TrainerControl": {
|
| 1419 |
+
"args": {
|
| 1420 |
+
"should_epoch_stop": false,
|
| 1421 |
+
"should_evaluate": false,
|
| 1422 |
+
"should_log": false,
|
| 1423 |
+
"should_save": true,
|
| 1424 |
+
"should_training_stop": false
|
| 1425 |
+
},
|
| 1426 |
+
"attributes": {}
|
| 1427 |
+
}
|
| 1428 |
+
},
|
| 1429 |
+
"total_flos": 5.934768918311731e+16,
|
| 1430 |
+
"train_batch_size": 4,
|
| 1431 |
+
"trial_name": null,
|
| 1432 |
+
"trial_params": null
|
| 1433 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-2000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:620747907b2130f2f54d285e74e3eb145407d8bba5652ffcf90771ba77206c41
|
| 3 |
+
size 5713
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "stanford-crfm/music-medium-800k",
|
| 3 |
+
"activation_function": "gelu_new",
|
| 4 |
+
"architectures": [
|
| 5 |
+
"GPT2LMHeadModel"
|
| 6 |
+
],
|
| 7 |
+
"attn_pdrop": 0.1,
|
| 8 |
+
"bos_token_id": 55025,
|
| 9 |
+
"embd_pdrop": 0.1,
|
| 10 |
+
"eos_token_id": 55025,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"layer_norm_epsilon": 1e-05,
|
| 13 |
+
"model_type": "gpt2",
|
| 14 |
+
"n_embd": 1024,
|
| 15 |
+
"n_head": 16,
|
| 16 |
+
"n_inner": null,
|
| 17 |
+
"n_layer": 24,
|
| 18 |
+
"n_positions": 1024,
|
| 19 |
+
"reorder_and_upcast_attn": false,
|
| 20 |
+
"resid_pdrop": 0.1,
|
| 21 |
+
"scale_attn_by_inverse_layer_idx": true,
|
| 22 |
+
"scale_attn_weights": true,
|
| 23 |
+
"summary_activation": null,
|
| 24 |
+
"summary_first_dropout": 0.1,
|
| 25 |
+
"summary_proj_to_labels": true,
|
| 26 |
+
"summary_type": "cls_index",
|
| 27 |
+
"summary_use_proj": true,
|
| 28 |
+
"torch_dtype": "float32",
|
| 29 |
+
"transformers_version": "4.49.0",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 55028
|
| 32 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 55025,
|
| 4 |
+
"eos_token_id": 55025,
|
| 5 |
+
"transformers_version": "4.49.0"
|
| 6 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37b8fbf2b682676eee61cd18f0ba95025451564f89382a681e1c7f13081705c1
|
| 3 |
+
size 2877913611
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1a5cf380d94e4d0a5f0ad6393d41f1aa967cd8e10c316ae78789be5a21c96cee
|
| 3 |
+
size 1438930739
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b570ada5d180275e6f12ba17d15851c2d1a968866b33f1d2fdc6435251f9685
|
| 3 |
+
size 14645
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0bb9c828a65fcb8fc82ccb373d736542631330622f28f2c20d7f42573db2816
|
| 3 |
+
size 1401
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,2133 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 9.346333853354134,
|
| 5 |
+
"eval_steps": 2000,
|
| 6 |
+
"global_step": 3000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.031201248049921998,
|
| 13 |
+
"grad_norm": 59933.58984375,
|
| 14 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 15 |
+
"loss": 13.4575,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.062402496099843996,
|
| 20 |
+
"grad_norm": 79273.921875,
|
| 21 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 22 |
+
"loss": 13.0486,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.093603744149766,
|
| 27 |
+
"grad_norm": 86346.1953125,
|
| 28 |
+
"learning_rate": 6.000000000000001e-07,
|
| 29 |
+
"loss": 11.9289,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12480499219968799,
|
| 34 |
+
"grad_norm": 68253.6484375,
|
| 35 |
+
"learning_rate": 8.000000000000001e-07,
|
| 36 |
+
"loss": 10.78,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15600624024961,
|
| 41 |
+
"grad_norm": 48220.171875,
|
| 42 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 43 |
+
"loss": 10.4398,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.187207488299532,
|
| 48 |
+
"grad_norm": 31435.85546875,
|
| 49 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 50 |
+
"loss": 9.2586,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21840873634945399,
|
| 55 |
+
"grad_norm": 11692.8125,
|
| 56 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 57 |
+
"loss": 8.4688,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24960998439937598,
|
| 62 |
+
"grad_norm": 33474.359375,
|
| 63 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 64 |
+
"loss": 7.8024,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.28081123244929795,
|
| 69 |
+
"grad_norm": 11187.919921875,
|
| 70 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 71 |
+
"loss": 7.5132,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.31201248049922,
|
| 76 |
+
"grad_norm": 6494.74609375,
|
| 77 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 78 |
+
"loss": 7.0462,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.34321372854914195,
|
| 83 |
+
"grad_norm": 4932.1884765625,
|
| 84 |
+
"learning_rate": 2.2e-06,
|
| 85 |
+
"loss": 6.6515,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.374414976599064,
|
| 90 |
+
"grad_norm": 14954.1064453125,
|
| 91 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 92 |
+
"loss": 6.352,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.40561622464898595,
|
| 97 |
+
"grad_norm": 8677.0498046875,
|
| 98 |
+
"learning_rate": 2.6e-06,
|
| 99 |
+
"loss": 5.9368,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.43681747269890797,
|
| 104 |
+
"grad_norm": 4330.2470703125,
|
| 105 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 106 |
+
"loss": 5.6135,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.46801872074882994,
|
| 111 |
+
"grad_norm": 4594.47216796875,
|
| 112 |
+
"learning_rate": 3e-06,
|
| 113 |
+
"loss": 5.3826,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.49921996879875197,
|
| 118 |
+
"grad_norm": 4907.52880859375,
|
| 119 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 120 |
+
"loss": 5.1467,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5304212168486739,
|
| 125 |
+
"grad_norm": 2622.095458984375,
|
| 126 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 127 |
+
"loss": 5.0465,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5616224648985959,
|
| 132 |
+
"grad_norm": 2206.215087890625,
|
| 133 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 134 |
+
"loss": 4.868,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.592823712948518,
|
| 139 |
+
"grad_norm": 1866.2569580078125,
|
| 140 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 141 |
+
"loss": 4.7356,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.62402496099844,
|
| 146 |
+
"grad_norm": 1359.1365966796875,
|
| 147 |
+
"learning_rate": 4.000000000000001e-06,
|
| 148 |
+
"loss": 4.4869,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6552262090483619,
|
| 153 |
+
"grad_norm": 929.0238647460938,
|
| 154 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 155 |
+
"loss": 4.2993,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6864274570982839,
|
| 160 |
+
"grad_norm": 814.4940795898438,
|
| 161 |
+
"learning_rate": 4.4e-06,
|
| 162 |
+
"loss": 4.0853,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.717628705148206,
|
| 167 |
+
"grad_norm": 561.3009033203125,
|
| 168 |
+
"learning_rate": 4.600000000000001e-06,
|
| 169 |
+
"loss": 3.9475,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.748829953198128,
|
| 174 |
+
"grad_norm": 595.680908203125,
|
| 175 |
+
"learning_rate": 4.800000000000001e-06,
|
| 176 |
+
"loss": 3.8025,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7800312012480499,
|
| 181 |
+
"grad_norm": 713.8980102539062,
|
| 182 |
+
"learning_rate": 5e-06,
|
| 183 |
+
"loss": 3.6036,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8112324492979719,
|
| 188 |
+
"grad_norm": 373.4093017578125,
|
| 189 |
+
"learning_rate": 5.2e-06,
|
| 190 |
+
"loss": 3.4541,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8424336973478939,
|
| 195 |
+
"grad_norm": 7371.640625,
|
| 196 |
+
"learning_rate": 5.400000000000001e-06,
|
| 197 |
+
"loss": 3.3097,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8736349453978159,
|
| 202 |
+
"grad_norm": 380.82952880859375,
|
| 203 |
+
"learning_rate": 5.600000000000001e-06,
|
| 204 |
+
"loss": 3.2375,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9048361934477379,
|
| 209 |
+
"grad_norm": 418.1549377441406,
|
| 210 |
+
"learning_rate": 5.8e-06,
|
| 211 |
+
"loss": 3.1552,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9360374414976599,
|
| 216 |
+
"grad_norm": 884.2369384765625,
|
| 217 |
+
"learning_rate": 6e-06,
|
| 218 |
+
"loss": 3.0261,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9672386895475819,
|
| 223 |
+
"grad_norm": 425.3170166015625,
|
| 224 |
+
"learning_rate": 6.200000000000001e-06,
|
| 225 |
+
"loss": 2.9638,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9984399375975039,
|
| 230 |
+
"grad_norm": 346.85650634765625,
|
| 231 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 232 |
+
"loss": 2.939,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0280811232449298,
|
| 237 |
+
"grad_norm": 607.9754028320312,
|
| 238 |
+
"learning_rate": 6.600000000000001e-06,
|
| 239 |
+
"loss": 2.8921,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0592823712948518,
|
| 244 |
+
"grad_norm": 73.27684020996094,
|
| 245 |
+
"learning_rate": 6.800000000000001e-06,
|
| 246 |
+
"loss": 2.802,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0904836193447738,
|
| 251 |
+
"grad_norm": 77.26670837402344,
|
| 252 |
+
"learning_rate": 7e-06,
|
| 253 |
+
"loss": 2.6704,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.1216848673946958,
|
| 258 |
+
"grad_norm": 46.717529296875,
|
| 259 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 260 |
+
"loss": 2.5633,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.1528861154446177,
|
| 265 |
+
"grad_norm": 73.30224609375,
|
| 266 |
+
"learning_rate": 7.4e-06,
|
| 267 |
+
"loss": 2.4874,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1840873634945397,
|
| 272 |
+
"grad_norm": 82.5370101928711,
|
| 273 |
+
"learning_rate": 7.600000000000001e-06,
|
| 274 |
+
"loss": 2.4168,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.2152886115444619,
|
| 279 |
+
"grad_norm": 56.439327239990234,
|
| 280 |
+
"learning_rate": 7.800000000000002e-06,
|
| 281 |
+
"loss": 2.2979,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2464898595943839,
|
| 286 |
+
"grad_norm": 34.27633285522461,
|
| 287 |
+
"learning_rate": 8.000000000000001e-06,
|
| 288 |
+
"loss": 2.2299,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2776911076443058,
|
| 293 |
+
"grad_norm": 24.567216873168945,
|
| 294 |
+
"learning_rate": 8.2e-06,
|
| 295 |
+
"loss": 2.1672,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3088923556942278,
|
| 300 |
+
"grad_norm": 28.32038688659668,
|
| 301 |
+
"learning_rate": 8.400000000000001e-06,
|
| 302 |
+
"loss": 2.1228,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3400936037441498,
|
| 307 |
+
"grad_norm": 47.05001449584961,
|
| 308 |
+
"learning_rate": 8.6e-06,
|
| 309 |
+
"loss": 2.0671,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3712948517940717,
|
| 314 |
+
"grad_norm": 59.189292907714844,
|
| 315 |
+
"learning_rate": 8.8e-06,
|
| 316 |
+
"loss": 1.9665,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4024960998439937,
|
| 321 |
+
"grad_norm": 62.23638916015625,
|
| 322 |
+
"learning_rate": 9e-06,
|
| 323 |
+
"loss": 1.9069,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4336973478939157,
|
| 328 |
+
"grad_norm": 20.689085006713867,
|
| 329 |
+
"learning_rate": 9.200000000000002e-06,
|
| 330 |
+
"loss": 1.8617,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4648985959438376,
|
| 335 |
+
"grad_norm": 21.1517333984375,
|
| 336 |
+
"learning_rate": 9.4e-06,
|
| 337 |
+
"loss": 1.8312,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4960998439937598,
|
| 342 |
+
"grad_norm": 11.305088996887207,
|
| 343 |
+
"learning_rate": 9.600000000000001e-06,
|
| 344 |
+
"loss": 1.7294,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5273010920436816,
|
| 349 |
+
"grad_norm": 10.452417373657227,
|
| 350 |
+
"learning_rate": 9.800000000000001e-06,
|
| 351 |
+
"loss": 1.641,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5585023400936038,
|
| 356 |
+
"grad_norm": 12.5308198928833,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 1.6129,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5897035881435257,
|
| 363 |
+
"grad_norm": 20.75917625427246,
|
| 364 |
+
"learning_rate": 9.999605221019082e-06,
|
| 365 |
+
"loss": 1.5093,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6209048361934477,
|
| 370 |
+
"grad_norm": 76.15613555908203,
|
| 371 |
+
"learning_rate": 9.9984209464165e-06,
|
| 372 |
+
"loss": 1.5462,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6521060842433697,
|
| 377 |
+
"grad_norm": 27.704729080200195,
|
| 378 |
+
"learning_rate": 9.996447363202947e-06,
|
| 379 |
+
"loss": 1.4907,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6833073322932917,
|
| 384 |
+
"grad_norm": 11.623372077941895,
|
| 385 |
+
"learning_rate": 9.99368478303009e-06,
|
| 386 |
+
"loss": 1.4383,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7145085803432139,
|
| 391 |
+
"grad_norm": 9.811054229736328,
|
| 392 |
+
"learning_rate": 9.990133642141359e-06,
|
| 393 |
+
"loss": 1.4416,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7457098283931356,
|
| 398 |
+
"grad_norm": 12.728073120117188,
|
| 399 |
+
"learning_rate": 9.98579450130307e-06,
|
| 400 |
+
"loss": 1.3726,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7769110764430578,
|
| 405 |
+
"grad_norm": 9.254298210144043,
|
| 406 |
+
"learning_rate": 9.980668045715864e-06,
|
| 407 |
+
"loss": 1.345,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8081123244929798,
|
| 412 |
+
"grad_norm": 19.775976181030273,
|
| 413 |
+
"learning_rate": 9.974755084906503e-06,
|
| 414 |
+
"loss": 1.3242,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8393135725429017,
|
| 419 |
+
"grad_norm": 6.770556926727295,
|
| 420 |
+
"learning_rate": 9.968056552600043e-06,
|
| 421 |
+
"loss": 1.292,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8705148205928237,
|
| 426 |
+
"grad_norm": 7.248960494995117,
|
| 427 |
+
"learning_rate": 9.960573506572391e-06,
|
| 428 |
+
"loss": 1.2693,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9017160686427457,
|
| 433 |
+
"grad_norm": 4.084126949310303,
|
| 434 |
+
"learning_rate": 9.952307128483257e-06,
|
| 435 |
+
"loss": 1.2154,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9329173166926679,
|
| 440 |
+
"grad_norm": 4.399320602416992,
|
| 441 |
+
"learning_rate": 9.94325872368957e-06,
|
| 442 |
+
"loss": 1.1869,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9641185647425896,
|
| 447 |
+
"grad_norm": 5.116558074951172,
|
| 448 |
+
"learning_rate": 9.93342972103934e-06,
|
| 449 |
+
"loss": 1.2168,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9953198127925118,
|
| 454 |
+
"grad_norm": 5.786620616912842,
|
| 455 |
+
"learning_rate": 9.922821672646028e-06,
|
| 456 |
+
"loss": 1.1734,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.0249609984399375,
|
| 461 |
+
"grad_norm": 15.972105979919434,
|
| 462 |
+
"learning_rate": 9.911436253643445e-06,
|
| 463 |
+
"loss": 1.1523,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0561622464898597,
|
| 468 |
+
"grad_norm": 4.64152193069458,
|
| 469 |
+
"learning_rate": 9.899275261921236e-06,
|
| 470 |
+
"loss": 1.1247,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0873634945397814,
|
| 475 |
+
"grad_norm": 7.009069442749023,
|
| 476 |
+
"learning_rate": 9.886340617840968e-06,
|
| 477 |
+
"loss": 1.0985,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.1185647425897036,
|
| 482 |
+
"grad_norm": 4.230483531951904,
|
| 483 |
+
"learning_rate": 9.872634363932887e-06,
|
| 484 |
+
"loss": 1.1014,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.149765990639626,
|
| 489 |
+
"grad_norm": 3.0746848583221436,
|
| 490 |
+
"learning_rate": 9.85815866457337e-06,
|
| 491 |
+
"loss": 1.0999,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.1809672386895476,
|
| 496 |
+
"grad_norm": 3.9389874935150146,
|
| 497 |
+
"learning_rate": 9.842915805643156e-06,
|
| 498 |
+
"loss": 1.0328,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.2121684867394698,
|
| 503 |
+
"grad_norm": 3.179098606109619,
|
| 504 |
+
"learning_rate": 9.82690819416637e-06,
|
| 505 |
+
"loss": 1.0448,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.2433697347893915,
|
| 510 |
+
"grad_norm": 3.1507132053375244,
|
| 511 |
+
"learning_rate": 9.81013835793043e-06,
|
| 512 |
+
"loss": 1.0195,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.2745709828393137,
|
| 517 |
+
"grad_norm": 3.040616273880005,
|
| 518 |
+
"learning_rate": 9.79260894508688e-06,
|
| 519 |
+
"loss": 0.9717,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.3057722308892354,
|
| 524 |
+
"grad_norm": 3.114654779434204,
|
| 525 |
+
"learning_rate": 9.774322723733216e-06,
|
| 526 |
+
"loss": 0.9784,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.3369734789391576,
|
| 531 |
+
"grad_norm": 4.01024866104126,
|
| 532 |
+
"learning_rate": 9.755282581475769e-06,
|
| 533 |
+
"loss": 0.9872,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.3681747269890794,
|
| 538 |
+
"grad_norm": 3.4513773918151855,
|
| 539 |
+
"learning_rate": 9.735491524973723e-06,
|
| 540 |
+
"loss": 0.9588,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3993759750390016,
|
| 545 |
+
"grad_norm": 3.086181879043579,
|
| 546 |
+
"learning_rate": 9.714952679464324e-06,
|
| 547 |
+
"loss": 0.9287,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4305772230889238,
|
| 552 |
+
"grad_norm": 2.971944570541382,
|
| 553 |
+
"learning_rate": 9.693669288269371e-06,
|
| 554 |
+
"loss": 0.9471,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.4617784711388455,
|
| 559 |
+
"grad_norm": 3.5111756324768066,
|
| 560 |
+
"learning_rate": 9.671644712283061e-06,
|
| 561 |
+
"loss": 0.9091,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4929797191887677,
|
| 566 |
+
"grad_norm": 3.4968149662017822,
|
| 567 |
+
"learning_rate": 9.648882429441258e-06,
|
| 568 |
+
"loss": 0.9133,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.5241809672386895,
|
| 573 |
+
"grad_norm": 2.7986488342285156,
|
| 574 |
+
"learning_rate": 9.62538603417229e-06,
|
| 575 |
+
"loss": 0.8803,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.5553822152886116,
|
| 580 |
+
"grad_norm": 2.635054349899292,
|
| 581 |
+
"learning_rate": 9.601159236829353e-06,
|
| 582 |
+
"loss": 0.904,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.5865834633385334,
|
| 587 |
+
"grad_norm": 2.8344554901123047,
|
| 588 |
+
"learning_rate": 9.576205863104588e-06,
|
| 589 |
+
"loss": 0.8549,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.6177847113884556,
|
| 594 |
+
"grad_norm": 29.804149627685547,
|
| 595 |
+
"learning_rate": 9.550529853424979e-06,
|
| 596 |
+
"loss": 0.8314,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.6489859594383773,
|
| 601 |
+
"grad_norm": 4.379691123962402,
|
| 602 |
+
"learning_rate": 9.524135262330098e-06,
|
| 603 |
+
"loss": 0.8286,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6801872074882995,
|
| 608 |
+
"grad_norm": 3.7554783821105957,
|
| 609 |
+
"learning_rate": 9.497026257831856e-06,
|
| 610 |
+
"loss": 0.8682,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.7113884555382217,
|
| 615 |
+
"grad_norm": 2.8485891819000244,
|
| 616 |
+
"learning_rate": 9.46920712075632e-06,
|
| 617 |
+
"loss": 0.8415,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.7425897035881435,
|
| 622 |
+
"grad_norm": 2.6579718589782715,
|
| 623 |
+
"learning_rate": 9.440682244067724e-06,
|
| 624 |
+
"loss": 0.8467,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7737909516380657,
|
| 629 |
+
"grad_norm": 3.369091272354126,
|
| 630 |
+
"learning_rate": 9.411456132174768e-06,
|
| 631 |
+
"loss": 0.8357,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8049921996879874,
|
| 636 |
+
"grad_norm": 2.6015613079071045,
|
| 637 |
+
"learning_rate": 9.381533400219319e-06,
|
| 638 |
+
"loss": 0.8351,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8361934477379096,
|
| 643 |
+
"grad_norm": 3.204615354537964,
|
| 644 |
+
"learning_rate": 9.35091877334763e-06,
|
| 645 |
+
"loss": 0.8059,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.8673946957878313,
|
| 650 |
+
"grad_norm": 3.4444634914398193,
|
| 651 |
+
"learning_rate": 9.319617085964177e-06,
|
| 652 |
+
"loss": 0.7631,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8985959438377535,
|
| 657 |
+
"grad_norm": 2.5985376834869385,
|
| 658 |
+
"learning_rate": 9.287633280968263e-06,
|
| 659 |
+
"loss": 0.788,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.9297971918876753,
|
| 664 |
+
"grad_norm": 2.7666327953338623,
|
| 665 |
+
"learning_rate": 9.25497240897346e-06,
|
| 666 |
+
"loss": 0.7521,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.9609984399375975,
|
| 671 |
+
"grad_norm": 2.8274083137512207,
|
| 672 |
+
"learning_rate": 9.221639627510076e-06,
|
| 673 |
+
"loss": 0.7878,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.9921996879875197,
|
| 678 |
+
"grad_norm": 2.522346019744873,
|
| 679 |
+
"learning_rate": 9.18764020021071e-06,
|
| 680 |
+
"loss": 0.7502,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0218408736349454,
|
| 685 |
+
"grad_norm": 2.3941545486450195,
|
| 686 |
+
"learning_rate": 9.152979495979064e-06,
|
| 687 |
+
"loss": 0.7451,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.0530421216848675,
|
| 692 |
+
"grad_norm": 3.3544349670410156,
|
| 693 |
+
"learning_rate": 9.117662988142138e-06,
|
| 694 |
+
"loss": 0.7193,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0842433697347893,
|
| 699 |
+
"grad_norm": 3.1165342330932617,
|
| 700 |
+
"learning_rate": 9.08169625358592e-06,
|
| 701 |
+
"loss": 0.7502,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.1154446177847115,
|
| 706 |
+
"grad_norm": 2.7650063037872314,
|
| 707 |
+
"learning_rate": 9.045084971874738e-06,
|
| 708 |
+
"loss": 0.7118,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.1466458658346332,
|
| 713 |
+
"grad_norm": 2.8753139972686768,
|
| 714 |
+
"learning_rate": 9.007834924354384e-06,
|
| 715 |
+
"loss": 0.7367,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.1778471138845554,
|
| 720 |
+
"grad_norm": 4.134742736816406,
|
| 721 |
+
"learning_rate": 8.969951993239177e-06,
|
| 722 |
+
"loss": 0.6477,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.209048361934477,
|
| 727 |
+
"grad_norm": 2.406000852584839,
|
| 728 |
+
"learning_rate": 8.931442160683094e-06,
|
| 729 |
+
"loss": 0.7034,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.2402496099843994,
|
| 734 |
+
"grad_norm": 2.5077903270721436,
|
| 735 |
+
"learning_rate": 8.892311507835118e-06,
|
| 736 |
+
"loss": 0.7009,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.2714508580343216,
|
| 741 |
+
"grad_norm": 2.7244954109191895,
|
| 742 |
+
"learning_rate": 8.852566213878947e-06,
|
| 743 |
+
"loss": 0.6858,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.3026521060842433,
|
| 748 |
+
"grad_norm": 2.5308499336242676,
|
| 749 |
+
"learning_rate": 8.81221255505724e-06,
|
| 750 |
+
"loss": 0.6717,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.3338533541341655,
|
| 755 |
+
"grad_norm": 2.368601083755493,
|
| 756 |
+
"learning_rate": 8.77125690368052e-06,
|
| 757 |
+
"loss": 0.6634,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.3650546021840873,
|
| 762 |
+
"grad_norm": 2.6542091369628906,
|
| 763 |
+
"learning_rate": 8.729705727120911e-06,
|
| 764 |
+
"loss": 0.6711,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.3962558502340094,
|
| 769 |
+
"grad_norm": 2.466643810272217,
|
| 770 |
+
"learning_rate": 8.68756558679087e-06,
|
| 771 |
+
"loss": 0.6418,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.427457098283931,
|
| 776 |
+
"grad_norm": 2.714127779006958,
|
| 777 |
+
"learning_rate": 8.644843137107058e-06,
|
| 778 |
+
"loss": 0.6524,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.4586583463338534,
|
| 783 |
+
"grad_norm": 2.508812427520752,
|
| 784 |
+
"learning_rate": 8.601545124439535e-06,
|
| 785 |
+
"loss": 0.6684,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.489859594383775,
|
| 790 |
+
"grad_norm": 2.2211365699768066,
|
| 791 |
+
"learning_rate": 8.557678386046427e-06,
|
| 792 |
+
"loss": 0.641,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.5210608424336973,
|
| 797 |
+
"grad_norm": 6.399374008178711,
|
| 798 |
+
"learning_rate": 8.513249848994248e-06,
|
| 799 |
+
"loss": 0.6831,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.5522620904836195,
|
| 804 |
+
"grad_norm": 2.5456409454345703,
|
| 805 |
+
"learning_rate": 8.468266529064025e-06,
|
| 806 |
+
"loss": 0.6274,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.5834633385335413,
|
| 811 |
+
"grad_norm": 2.545017719268799,
|
| 812 |
+
"learning_rate": 8.422735529643445e-06,
|
| 813 |
+
"loss": 0.6361,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.6146645865834635,
|
| 818 |
+
"grad_norm": 2.749608278274536,
|
| 819 |
+
"learning_rate": 8.376664040605122e-06,
|
| 820 |
+
"loss": 0.6233,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.645865834633385,
|
| 825 |
+
"grad_norm": 2.785017728805542,
|
| 826 |
+
"learning_rate": 8.33005933717126e-06,
|
| 827 |
+
"loss": 0.6125,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.6770670826833074,
|
| 832 |
+
"grad_norm": 2.522922992706299,
|
| 833 |
+
"learning_rate": 8.282928778764783e-06,
|
| 834 |
+
"loss": 0.5696,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.7082683307332296,
|
| 839 |
+
"grad_norm": 2.222412347793579,
|
| 840 |
+
"learning_rate": 8.235279807847223e-06,
|
| 841 |
+
"loss": 0.6293,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.7394695787831513,
|
| 846 |
+
"grad_norm": 2.597766399383545,
|
| 847 |
+
"learning_rate": 8.18711994874345e-06,
|
| 848 |
+
"loss": 0.6062,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.770670826833073,
|
| 853 |
+
"grad_norm": 2.3702147006988525,
|
| 854 |
+
"learning_rate": 8.138456806453503e-06,
|
| 855 |
+
"loss": 0.5789,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.8018720748829953,
|
| 860 |
+
"grad_norm": 4.38936185836792,
|
| 861 |
+
"learning_rate": 8.089298065451673e-06,
|
| 862 |
+
"loss": 0.6036,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.8330733229329175,
|
| 867 |
+
"grad_norm": 2.9343149662017822,
|
| 868 |
+
"learning_rate": 8.039651488473028e-06,
|
| 869 |
+
"loss": 0.5832,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.864274570982839,
|
| 874 |
+
"grad_norm": 2.2963383197784424,
|
| 875 |
+
"learning_rate": 7.989524915287595e-06,
|
| 876 |
+
"loss": 0.5975,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.8954758190327614,
|
| 881 |
+
"grad_norm": 7.433852672576904,
|
| 882 |
+
"learning_rate": 7.938926261462366e-06,
|
| 883 |
+
"loss": 0.594,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.926677067082683,
|
| 888 |
+
"grad_norm": 2.5729384422302246,
|
| 889 |
+
"learning_rate": 7.887863517111337e-06,
|
| 890 |
+
"loss": 0.5937,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.9578783151326054,
|
| 895 |
+
"grad_norm": 2.6089067459106445,
|
| 896 |
+
"learning_rate": 7.836344745633785e-06,
|
| 897 |
+
"loss": 0.6126,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.9890795631825275,
|
| 902 |
+
"grad_norm": 2.289501667022705,
|
| 903 |
+
"learning_rate": 7.78437808244094e-06,
|
| 904 |
+
"loss": 0.5892,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.018720748829953,
|
| 909 |
+
"grad_norm": 2.9094228744506836,
|
| 910 |
+
"learning_rate": 7.731971733671347e-06,
|
| 911 |
+
"loss": 0.5797,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.049921996879875,
|
| 916 |
+
"grad_norm": 2.0647175312042236,
|
| 917 |
+
"learning_rate": 7.679133974894984e-06,
|
| 918 |
+
"loss": 0.5628,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.081123244929797,
|
| 923 |
+
"grad_norm": 2.374997854232788,
|
| 924 |
+
"learning_rate": 7.6258731498064796e-06,
|
| 925 |
+
"loss": 0.5923,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.112324492979719,
|
| 930 |
+
"grad_norm": 2.2724082469940186,
|
| 931 |
+
"learning_rate": 7.572197668907533e-06,
|
| 932 |
+
"loss": 0.5574,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.1435257410296416,
|
| 937 |
+
"grad_norm": 2.1879308223724365,
|
| 938 |
+
"learning_rate": 7.518116008178805e-06,
|
| 939 |
+
"loss": 0.5568,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.174726989079563,
|
| 944 |
+
"grad_norm": 2.5050461292266846,
|
| 945 |
+
"learning_rate": 7.463636707741458e-06,
|
| 946 |
+
"loss": 0.526,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.205928237129485,
|
| 951 |
+
"grad_norm": 2.2555339336395264,
|
| 952 |
+
"learning_rate": 7.408768370508577e-06,
|
| 953 |
+
"loss": 0.5302,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.237129485179407,
|
| 958 |
+
"grad_norm": 2.5363669395446777,
|
| 959 |
+
"learning_rate": 7.353519660826665e-06,
|
| 960 |
+
"loss": 0.5485,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.268330733229329,
|
| 965 |
+
"grad_norm": 2.200976848602295,
|
| 966 |
+
"learning_rate": 7.297899303107441e-06,
|
| 967 |
+
"loss": 0.5301,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.299531981279252,
|
| 972 |
+
"grad_norm": 2.2207038402557373,
|
| 973 |
+
"learning_rate": 7.241916080450163e-06,
|
| 974 |
+
"loss": 0.51,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.330733229329173,
|
| 979 |
+
"grad_norm": 2.300055980682373,
|
| 980 |
+
"learning_rate": 7.185578833254665e-06,
|
| 981 |
+
"loss": 0.5216,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.361934477379095,
|
| 986 |
+
"grad_norm": 2.3568477630615234,
|
| 987 |
+
"learning_rate": 7.128896457825364e-06,
|
| 988 |
+
"loss": 0.545,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.393135725429017,
|
| 993 |
+
"grad_norm": 2.1130895614624023,
|
| 994 |
+
"learning_rate": 7.071877904966422e-06,
|
| 995 |
+
"loss": 0.5231,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.4243369734789395,
|
| 1000 |
+
"grad_norm": 2.3051438331604004,
|
| 1001 |
+
"learning_rate": 7.014532178568314e-06,
|
| 1002 |
+
"loss": 0.5072,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.455538221528861,
|
| 1007 |
+
"grad_norm": 2.8409862518310547,
|
| 1008 |
+
"learning_rate": 6.9568683341860135e-06,
|
| 1009 |
+
"loss": 0.4753,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.486739469578783,
|
| 1014 |
+
"grad_norm": 2.3482327461242676,
|
| 1015 |
+
"learning_rate": 6.898895477609007e-06,
|
| 1016 |
+
"loss": 0.5158,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.517940717628705,
|
| 1021 |
+
"grad_norm": 2.4122118949890137,
|
| 1022 |
+
"learning_rate": 6.840622763423391e-06,
|
| 1023 |
+
"loss": 0.518,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.549141965678627,
|
| 1028 |
+
"grad_norm": 2.0749752521514893,
|
| 1029 |
+
"learning_rate": 6.782059393566254e-06,
|
| 1030 |
+
"loss": 0.4895,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.580343213728549,
|
| 1035 |
+
"grad_norm": 2.186260461807251,
|
| 1036 |
+
"learning_rate": 6.723214615872585e-06,
|
| 1037 |
+
"loss": 0.4872,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.611544461778471,
|
| 1042 |
+
"grad_norm": 2.3305256366729736,
|
| 1043 |
+
"learning_rate": 6.664097722614934e-06,
|
| 1044 |
+
"loss": 0.4935,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.642745709828393,
|
| 1049 |
+
"grad_norm": 2.2758896350860596,
|
| 1050 |
+
"learning_rate": 6.604718049036047e-06,
|
| 1051 |
+
"loss": 0.4942,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.673946957878315,
|
| 1056 |
+
"grad_norm": 3.157252311706543,
|
| 1057 |
+
"learning_rate": 6.545084971874738e-06,
|
| 1058 |
+
"loss": 0.4939,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.7051482059282375,
|
| 1063 |
+
"grad_norm": 2.2830564975738525,
|
| 1064 |
+
"learning_rate": 6.485207907885175e-06,
|
| 1065 |
+
"loss": 0.5057,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.736349453978159,
|
| 1070 |
+
"grad_norm": 2.0698862075805664,
|
| 1071 |
+
"learning_rate": 6.425096312349881e-06,
|
| 1072 |
+
"loss": 0.4922,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.767550702028081,
|
| 1077 |
+
"grad_norm": 2.110299587249756,
|
| 1078 |
+
"learning_rate": 6.364759677586627e-06,
|
| 1079 |
+
"loss": 0.4821,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.798751950078003,
|
| 1084 |
+
"grad_norm": 2.074535369873047,
|
| 1085 |
+
"learning_rate": 6.304207531449486e-06,
|
| 1086 |
+
"loss": 0.4652,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.829953198127925,
|
| 1091 |
+
"grad_norm": 2.317800998687744,
|
| 1092 |
+
"learning_rate": 6.243449435824276e-06,
|
| 1093 |
+
"loss": 0.4867,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.8611544461778475,
|
| 1098 |
+
"grad_norm": 2.319498062133789,
|
| 1099 |
+
"learning_rate": 6.182494985118625e-06,
|
| 1100 |
+
"loss": 0.4316,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.892355694227769,
|
| 1105 |
+
"grad_norm": 2.2666006088256836,
|
| 1106 |
+
"learning_rate": 6.121353804746907e-06,
|
| 1107 |
+
"loss": 0.466,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.923556942277691,
|
| 1112 |
+
"grad_norm": 2.0361239910125732,
|
| 1113 |
+
"learning_rate": 6.060035549610275e-06,
|
| 1114 |
+
"loss": 0.4615,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.954758190327613,
|
| 1119 |
+
"grad_norm": 2.0396602153778076,
|
| 1120 |
+
"learning_rate": 5.9985499025720354e-06,
|
| 1121 |
+
"loss": 0.4672,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.985959438377535,
|
| 1126 |
+
"grad_norm": 1.9205994606018066,
|
| 1127 |
+
"learning_rate": 5.936906572928625e-06,
|
| 1128 |
+
"loss": 0.4197,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 5.015600624024961,
|
| 1133 |
+
"grad_norm": 2.4138858318328857,
|
| 1134 |
+
"learning_rate": 5.8751152948763815e-06,
|
| 1135 |
+
"loss": 0.4384,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 5.046801872074883,
|
| 1140 |
+
"grad_norm": 2.2438735961914062,
|
| 1141 |
+
"learning_rate": 5.813185825974419e-06,
|
| 1142 |
+
"loss": 0.4392,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 5.078003120124805,
|
| 1147 |
+
"grad_norm": 2.7369537353515625,
|
| 1148 |
+
"learning_rate": 5.751127945603786e-06,
|
| 1149 |
+
"loss": 0.4654,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 5.109204368174727,
|
| 1154 |
+
"grad_norm": 2.676726818084717,
|
| 1155 |
+
"learning_rate": 5.68895145342319e-06,
|
| 1156 |
+
"loss": 0.4572,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 5.140405616224649,
|
| 1161 |
+
"grad_norm": 2.2318201065063477,
|
| 1162 |
+
"learning_rate": 5.626666167821522e-06,
|
| 1163 |
+
"loss": 0.4801,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 5.171606864274571,
|
| 1168 |
+
"grad_norm": 2.003933906555176,
|
| 1169 |
+
"learning_rate": 5.5642819243674085e-06,
|
| 1170 |
+
"loss": 0.4197,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 5.202808112324493,
|
| 1175 |
+
"grad_norm": 1.9387847185134888,
|
| 1176 |
+
"learning_rate": 5.5018085742560745e-06,
|
| 1177 |
+
"loss": 0.4372,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 5.234009360374415,
|
| 1182 |
+
"grad_norm": 2.0050766468048096,
|
| 1183 |
+
"learning_rate": 5.439255982753717e-06,
|
| 1184 |
+
"loss": 0.4553,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 5.265210608424337,
|
| 1189 |
+
"grad_norm": 2.092911720275879,
|
| 1190 |
+
"learning_rate": 5.376634027639664e-06,
|
| 1191 |
+
"loss": 0.4383,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 5.296411856474259,
|
| 1196 |
+
"grad_norm": 1.9465577602386475,
|
| 1197 |
+
"learning_rate": 5.3139525976465675e-06,
|
| 1198 |
+
"loss": 0.4549,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 5.327613104524181,
|
| 1203 |
+
"grad_norm": 2.0983693599700928,
|
| 1204 |
+
"learning_rate": 5.251221590898848e-06,
|
| 1205 |
+
"loss": 0.4639,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 5.358814352574103,
|
| 1210 |
+
"grad_norm": 1.85776686668396,
|
| 1211 |
+
"learning_rate": 5.188450913349674e-06,
|
| 1212 |
+
"loss": 0.4129,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 5.390015600624025,
|
| 1217 |
+
"grad_norm": 1.9787360429763794,
|
| 1218 |
+
"learning_rate": 5.1256504772166885e-06,
|
| 1219 |
+
"loss": 0.4416,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 5.4212168486739465,
|
| 1224 |
+
"grad_norm": 2.25137996673584,
|
| 1225 |
+
"learning_rate": 5.062830199416764e-06,
|
| 1226 |
+
"loss": 0.4325,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 5.452418096723869,
|
| 1231 |
+
"grad_norm": 2.2280070781707764,
|
| 1232 |
+
"learning_rate": 5e-06,
|
| 1233 |
+
"loss": 0.4061,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 5.483619344773791,
|
| 1238 |
+
"grad_norm": 1.9798814058303833,
|
| 1239 |
+
"learning_rate": 4.937169800583237e-06,
|
| 1240 |
+
"loss": 0.4398,
|
| 1241 |
+
"step": 1760
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 5.514820592823713,
|
| 1245 |
+
"grad_norm": 1.8920650482177734,
|
| 1246 |
+
"learning_rate": 4.874349522783313e-06,
|
| 1247 |
+
"loss": 0.4009,
|
| 1248 |
+
"step": 1770
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 5.546021840873635,
|
| 1252 |
+
"grad_norm": 2.217628240585327,
|
| 1253 |
+
"learning_rate": 4.811549086650327e-06,
|
| 1254 |
+
"loss": 0.4084,
|
| 1255 |
+
"step": 1780
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 5.577223088923557,
|
| 1259 |
+
"grad_norm": 2.321141481399536,
|
| 1260 |
+
"learning_rate": 4.748778409101153e-06,
|
| 1261 |
+
"loss": 0.4229,
|
| 1262 |
+
"step": 1790
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 5.608424336973479,
|
| 1266 |
+
"grad_norm": 2.1298415660858154,
|
| 1267 |
+
"learning_rate": 4.686047402353433e-06,
|
| 1268 |
+
"loss": 0.3925,
|
| 1269 |
+
"step": 1800
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 5.639625585023401,
|
| 1273 |
+
"grad_norm": 1.9593074321746826,
|
| 1274 |
+
"learning_rate": 4.6233659723603374e-06,
|
| 1275 |
+
"loss": 0.4292,
|
| 1276 |
+
"step": 1810
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 5.670826833073323,
|
| 1280 |
+
"grad_norm": 2.3608038425445557,
|
| 1281 |
+
"learning_rate": 4.560744017246284e-06,
|
| 1282 |
+
"loss": 0.3864,
|
| 1283 |
+
"step": 1820
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 5.702028081123245,
|
| 1287 |
+
"grad_norm": 1.8943296670913696,
|
| 1288 |
+
"learning_rate": 4.4981914257439254e-06,
|
| 1289 |
+
"loss": 0.4034,
|
| 1290 |
+
"step": 1830
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 5.733229329173167,
|
| 1294 |
+
"grad_norm": 2.0361130237579346,
|
| 1295 |
+
"learning_rate": 4.4357180756325915e-06,
|
| 1296 |
+
"loss": 0.3829,
|
| 1297 |
+
"step": 1840
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 5.764430577223089,
|
| 1301 |
+
"grad_norm": 1.977895975112915,
|
| 1302 |
+
"learning_rate": 4.373333832178478e-06,
|
| 1303 |
+
"loss": 0.4345,
|
| 1304 |
+
"step": 1850
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 5.795631825273011,
|
| 1308 |
+
"grad_norm": 2.087878465652466,
|
| 1309 |
+
"learning_rate": 4.31104854657681e-06,
|
| 1310 |
+
"loss": 0.4115,
|
| 1311 |
+
"step": 1860
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 5.826833073322933,
|
| 1315 |
+
"grad_norm": 2.121720314025879,
|
| 1316 |
+
"learning_rate": 4.248872054396215e-06,
|
| 1317 |
+
"loss": 0.3975,
|
| 1318 |
+
"step": 1870
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 5.858034321372855,
|
| 1322 |
+
"grad_norm": 2.06178879737854,
|
| 1323 |
+
"learning_rate": 4.186814174025582e-06,
|
| 1324 |
+
"loss": 0.4194,
|
| 1325 |
+
"step": 1880
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 5.889235569422777,
|
| 1329 |
+
"grad_norm": 1.8563209772109985,
|
| 1330 |
+
"learning_rate": 4.124884705123619e-06,
|
| 1331 |
+
"loss": 0.4003,
|
| 1332 |
+
"step": 1890
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 5.920436817472699,
|
| 1336 |
+
"grad_norm": 2.121776580810547,
|
| 1337 |
+
"learning_rate": 4.063093427071376e-06,
|
| 1338 |
+
"loss": 0.4409,
|
| 1339 |
+
"step": 1900
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 5.951638065522621,
|
| 1343 |
+
"grad_norm": 2.0366406440734863,
|
| 1344 |
+
"learning_rate": 4.001450097427965e-06,
|
| 1345 |
+
"loss": 0.3787,
|
| 1346 |
+
"step": 1910
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 5.982839313572543,
|
| 1350 |
+
"grad_norm": 1.861203670501709,
|
| 1351 |
+
"learning_rate": 3.939964450389728e-06,
|
| 1352 |
+
"loss": 0.4429,
|
| 1353 |
+
"step": 1920
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 6.0124804992199685,
|
| 1357 |
+
"grad_norm": 2.0258266925811768,
|
| 1358 |
+
"learning_rate": 3.8786461952530955e-06,
|
| 1359 |
+
"loss": 0.3855,
|
| 1360 |
+
"step": 1930
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 6.043681747269891,
|
| 1364 |
+
"grad_norm": 1.9031906127929688,
|
| 1365 |
+
"learning_rate": 3.817505014881378e-06,
|
| 1366 |
+
"loss": 0.3763,
|
| 1367 |
+
"step": 1940
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 6.074882995319813,
|
| 1371 |
+
"grad_norm": 2.047241449356079,
|
| 1372 |
+
"learning_rate": 3.756550564175727e-06,
|
| 1373 |
+
"loss": 0.3668,
|
| 1374 |
+
"step": 1950
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 6.106084243369735,
|
| 1378 |
+
"grad_norm": 2.1982035636901855,
|
| 1379 |
+
"learning_rate": 3.695792468550517e-06,
|
| 1380 |
+
"loss": 0.3892,
|
| 1381 |
+
"step": 1960
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 6.137285491419656,
|
| 1385 |
+
"grad_norm": 1.949191689491272,
|
| 1386 |
+
"learning_rate": 3.635240322413375e-06,
|
| 1387 |
+
"loss": 0.418,
|
| 1388 |
+
"step": 1970
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 6.168486739469579,
|
| 1392 |
+
"grad_norm": 2.028028726577759,
|
| 1393 |
+
"learning_rate": 3.5749036876501196e-06,
|
| 1394 |
+
"loss": 0.3972,
|
| 1395 |
+
"step": 1980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 6.199687987519501,
|
| 1399 |
+
"grad_norm": 2.0211682319641113,
|
| 1400 |
+
"learning_rate": 3.5147920921148267e-06,
|
| 1401 |
+
"loss": 0.4032,
|
| 1402 |
+
"step": 1990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 6.230889235569423,
|
| 1406 |
+
"grad_norm": 2.1258091926574707,
|
| 1407 |
+
"learning_rate": 3.4549150281252635e-06,
|
| 1408 |
+
"loss": 0.3885,
|
| 1409 |
+
"step": 2000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 6.262090483619345,
|
| 1413 |
+
"grad_norm": 1.9776920080184937,
|
| 1414 |
+
"learning_rate": 3.3952819509639534e-06,
|
| 1415 |
+
"loss": 0.4126,
|
| 1416 |
+
"step": 2010
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 6.2932917316692665,
|
| 1420 |
+
"grad_norm": 2.2959957122802734,
|
| 1421 |
+
"learning_rate": 3.3359022773850673e-06,
|
| 1422 |
+
"loss": 0.3777,
|
| 1423 |
+
"step": 2020
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 6.324492979719189,
|
| 1427 |
+
"grad_norm": 1.9055465459823608,
|
| 1428 |
+
"learning_rate": 3.2767853841274154e-06,
|
| 1429 |
+
"loss": 0.3838,
|
| 1430 |
+
"step": 2030
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 6.355694227769111,
|
| 1434 |
+
"grad_norm": 2.0175342559814453,
|
| 1435 |
+
"learning_rate": 3.217940606433747e-06,
|
| 1436 |
+
"loss": 0.4048,
|
| 1437 |
+
"step": 2040
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 6.386895475819033,
|
| 1441 |
+
"grad_norm": 1.966408371925354,
|
| 1442 |
+
"learning_rate": 3.1593772365766107e-06,
|
| 1443 |
+
"loss": 0.3758,
|
| 1444 |
+
"step": 2050
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 6.418096723868954,
|
| 1448 |
+
"grad_norm": 1.8205311298370361,
|
| 1449 |
+
"learning_rate": 3.1011045223909954e-06,
|
| 1450 |
+
"loss": 0.3415,
|
| 1451 |
+
"step": 2060
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 6.4492979719188765,
|
| 1455 |
+
"grad_norm": 1.9581196308135986,
|
| 1456 |
+
"learning_rate": 3.043131665813988e-06,
|
| 1457 |
+
"loss": 0.3788,
|
| 1458 |
+
"step": 2070
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 6.480499219968799,
|
| 1462 |
+
"grad_norm": 1.9435299634933472,
|
| 1463 |
+
"learning_rate": 2.9854678214316875e-06,
|
| 1464 |
+
"loss": 0.3735,
|
| 1465 |
+
"step": 2080
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 6.511700468018721,
|
| 1469 |
+
"grad_norm": 2.035621404647827,
|
| 1470 |
+
"learning_rate": 2.92812209503358e-06,
|
| 1471 |
+
"loss": 0.359,
|
| 1472 |
+
"step": 2090
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 6.542901716068643,
|
| 1476 |
+
"grad_norm": 1.8920469284057617,
|
| 1477 |
+
"learning_rate": 2.871103542174637e-06,
|
| 1478 |
+
"loss": 0.3819,
|
| 1479 |
+
"step": 2100
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 6.574102964118564,
|
| 1483 |
+
"grad_norm": 1.837669014930725,
|
| 1484 |
+
"learning_rate": 2.814421166745337e-06,
|
| 1485 |
+
"loss": 0.3854,
|
| 1486 |
+
"step": 2110
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 6.605304212168487,
|
| 1490 |
+
"grad_norm": 1.994737148284912,
|
| 1491 |
+
"learning_rate": 2.7580839195498397e-06,
|
| 1492 |
+
"loss": 0.4095,
|
| 1493 |
+
"step": 2120
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 6.636505460218409,
|
| 1497 |
+
"grad_norm": 1.936189889907837,
|
| 1498 |
+
"learning_rate": 2.7021006968925613e-06,
|
| 1499 |
+
"loss": 0.3713,
|
| 1500 |
+
"step": 2130
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 6.667706708268331,
|
| 1504 |
+
"grad_norm": 1.893553376197815,
|
| 1505 |
+
"learning_rate": 2.646480339173337e-06,
|
| 1506 |
+
"loss": 0.3782,
|
| 1507 |
+
"step": 2140
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 6.698907956318253,
|
| 1511 |
+
"grad_norm": 2.00282883644104,
|
| 1512 |
+
"learning_rate": 2.5912316294914232e-06,
|
| 1513 |
+
"loss": 0.4226,
|
| 1514 |
+
"step": 2150
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 6.7301092043681745,
|
| 1518 |
+
"grad_norm": 1.7351257801055908,
|
| 1519 |
+
"learning_rate": 2.536363292258543e-06,
|
| 1520 |
+
"loss": 0.3535,
|
| 1521 |
+
"step": 2160
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 6.761310452418097,
|
| 1525 |
+
"grad_norm": 1.865121841430664,
|
| 1526 |
+
"learning_rate": 2.4818839918211963e-06,
|
| 1527 |
+
"loss": 0.3649,
|
| 1528 |
+
"step": 2170
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 6.792511700468019,
|
| 1532 |
+
"grad_norm": 2.3301901817321777,
|
| 1533 |
+
"learning_rate": 2.4278023310924676e-06,
|
| 1534 |
+
"loss": 0.3822,
|
| 1535 |
+
"step": 2180
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 6.823712948517941,
|
| 1539 |
+
"grad_norm": 1.752894401550293,
|
| 1540 |
+
"learning_rate": 2.3741268501935212e-06,
|
| 1541 |
+
"loss": 0.3451,
|
| 1542 |
+
"step": 2190
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 6.854914196567862,
|
| 1546 |
+
"grad_norm": 1.6319270133972168,
|
| 1547 |
+
"learning_rate": 2.320866025105016e-06,
|
| 1548 |
+
"loss": 0.3911,
|
| 1549 |
+
"step": 2200
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 6.886115444617785,
|
| 1553 |
+
"grad_norm": 1.9505712985992432,
|
| 1554 |
+
"learning_rate": 2.268028266328655e-06,
|
| 1555 |
+
"loss": 0.3711,
|
| 1556 |
+
"step": 2210
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 6.917316692667707,
|
| 1560 |
+
"grad_norm": 1.9390125274658203,
|
| 1561 |
+
"learning_rate": 2.2156219175590623e-06,
|
| 1562 |
+
"loss": 0.3637,
|
| 1563 |
+
"step": 2220
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 6.948517940717629,
|
| 1567 |
+
"grad_norm": 1.9121036529541016,
|
| 1568 |
+
"learning_rate": 2.1636552543662187e-06,
|
| 1569 |
+
"loss": 0.3886,
|
| 1570 |
+
"step": 2230
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 6.97971918876755,
|
| 1574 |
+
"grad_norm": 1.8223578929901123,
|
| 1575 |
+
"learning_rate": 2.112136482888663e-06,
|
| 1576 |
+
"loss": 0.3537,
|
| 1577 |
+
"step": 2240
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.009360374414976,
|
| 1581 |
+
"grad_norm": 2.0081751346588135,
|
| 1582 |
+
"learning_rate": 2.061073738537635e-06,
|
| 1583 |
+
"loss": 0.3812,
|
| 1584 |
+
"step": 2250
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.040561622464899,
|
| 1588 |
+
"grad_norm": 1.7976573705673218,
|
| 1589 |
+
"learning_rate": 2.0104750847124075e-06,
|
| 1590 |
+
"loss": 0.3525,
|
| 1591 |
+
"step": 2260
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.071762870514821,
|
| 1595 |
+
"grad_norm": 2.144692897796631,
|
| 1596 |
+
"learning_rate": 1.9603485115269743e-06,
|
| 1597 |
+
"loss": 0.3679,
|
| 1598 |
+
"step": 2270
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.102964118564743,
|
| 1602 |
+
"grad_norm": 2.275657892227173,
|
| 1603 |
+
"learning_rate": 1.910701934548329e-06,
|
| 1604 |
+
"loss": 0.3469,
|
| 1605 |
+
"step": 2280
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.134165366614664,
|
| 1609 |
+
"grad_norm": 1.7523900270462036,
|
| 1610 |
+
"learning_rate": 1.8615431935464984e-06,
|
| 1611 |
+
"loss": 0.3341,
|
| 1612 |
+
"step": 2290
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.1653666146645865,
|
| 1616 |
+
"grad_norm": 1.955221176147461,
|
| 1617 |
+
"learning_rate": 1.8128800512565514e-06,
|
| 1618 |
+
"loss": 0.369,
|
| 1619 |
+
"step": 2300
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.196567862714509,
|
| 1623 |
+
"grad_norm": 1.8861024379730225,
|
| 1624 |
+
"learning_rate": 1.7647201921527802e-06,
|
| 1625 |
+
"loss": 0.3345,
|
| 1626 |
+
"step": 2310
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.227769110764431,
|
| 1630 |
+
"grad_norm": 1.7162634134292603,
|
| 1631 |
+
"learning_rate": 1.7170712212352187e-06,
|
| 1632 |
+
"loss": 0.355,
|
| 1633 |
+
"step": 2320
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 7.258970358814352,
|
| 1637 |
+
"grad_norm": 1.7793327569961548,
|
| 1638 |
+
"learning_rate": 1.6699406628287423e-06,
|
| 1639 |
+
"loss": 0.3672,
|
| 1640 |
+
"step": 2330
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 7.290171606864274,
|
| 1644 |
+
"grad_norm": 1.953660488128662,
|
| 1645 |
+
"learning_rate": 1.6233359593948784e-06,
|
| 1646 |
+
"loss": 0.3709,
|
| 1647 |
+
"step": 2340
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 7.3213728549141965,
|
| 1651 |
+
"grad_norm": 1.7876272201538086,
|
| 1652 |
+
"learning_rate": 1.5772644703565564e-06,
|
| 1653 |
+
"loss": 0.3543,
|
| 1654 |
+
"step": 2350
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 7.352574102964119,
|
| 1658 |
+
"grad_norm": 1.4949634075164795,
|
| 1659 |
+
"learning_rate": 1.5317334709359767e-06,
|
| 1660 |
+
"loss": 0.3341,
|
| 1661 |
+
"step": 2360
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 7.383775351014041,
|
| 1665 |
+
"grad_norm": 1.831666350364685,
|
| 1666 |
+
"learning_rate": 1.4867501510057548e-06,
|
| 1667 |
+
"loss": 0.3512,
|
| 1668 |
+
"step": 2370
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 7.414976599063962,
|
| 1672 |
+
"grad_norm": 1.8930827379226685,
|
| 1673 |
+
"learning_rate": 1.4423216139535735e-06,
|
| 1674 |
+
"loss": 0.3585,
|
| 1675 |
+
"step": 2380
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 7.446177847113884,
|
| 1679 |
+
"grad_norm": 1.9929951429367065,
|
| 1680 |
+
"learning_rate": 1.3984548755604655e-06,
|
| 1681 |
+
"loss": 0.3445,
|
| 1682 |
+
"step": 2390
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 7.477379095163807,
|
| 1686 |
+
"grad_norm": 1.9674276113510132,
|
| 1687 |
+
"learning_rate": 1.3551568628929434e-06,
|
| 1688 |
+
"loss": 0.3997,
|
| 1689 |
+
"step": 2400
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 7.508580343213729,
|
| 1693 |
+
"grad_norm": 1.8136045932769775,
|
| 1694 |
+
"learning_rate": 1.312434413209131e-06,
|
| 1695 |
+
"loss": 0.3727,
|
| 1696 |
+
"step": 2410
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 7.539781591263651,
|
| 1700 |
+
"grad_norm": 1.7950389385223389,
|
| 1701 |
+
"learning_rate": 1.2702942728790897e-06,
|
| 1702 |
+
"loss": 0.3766,
|
| 1703 |
+
"step": 2420
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 7.570982839313572,
|
| 1707 |
+
"grad_norm": 1.7054840326309204,
|
| 1708 |
+
"learning_rate": 1.2287430963194807e-06,
|
| 1709 |
+
"loss": 0.3847,
|
| 1710 |
+
"step": 2430
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 7.6021840873634945,
|
| 1714 |
+
"grad_norm": 1.888563871383667,
|
| 1715 |
+
"learning_rate": 1.18778744494276e-06,
|
| 1716 |
+
"loss": 0.378,
|
| 1717 |
+
"step": 2440
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 7.633385335413417,
|
| 1721 |
+
"grad_norm": 1.9313362836837769,
|
| 1722 |
+
"learning_rate": 1.1474337861210543e-06,
|
| 1723 |
+
"loss": 0.3721,
|
| 1724 |
+
"step": 2450
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 7.664586583463339,
|
| 1728 |
+
"grad_norm": 1.867435097694397,
|
| 1729 |
+
"learning_rate": 1.1076884921648834e-06,
|
| 1730 |
+
"loss": 0.3661,
|
| 1731 |
+
"step": 2460
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 7.69578783151326,
|
| 1735 |
+
"grad_norm": 1.7566686868667603,
|
| 1736 |
+
"learning_rate": 1.0685578393169054e-06,
|
| 1737 |
+
"loss": 0.4059,
|
| 1738 |
+
"step": 2470
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 7.726989079563182,
|
| 1742 |
+
"grad_norm": 2.269183874130249,
|
| 1743 |
+
"learning_rate": 1.0300480067608232e-06,
|
| 1744 |
+
"loss": 0.3633,
|
| 1745 |
+
"step": 2480
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 7.758190327613105,
|
| 1749 |
+
"grad_norm": 1.9920068979263306,
|
| 1750 |
+
"learning_rate": 9.921650756456164e-07,
|
| 1751 |
+
"loss": 0.3741,
|
| 1752 |
+
"step": 2490
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 7.789391575663027,
|
| 1756 |
+
"grad_norm": 1.7317137718200684,
|
| 1757 |
+
"learning_rate": 9.549150281252633e-07,
|
| 1758 |
+
"loss": 0.3528,
|
| 1759 |
+
"step": 2500
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 7.820592823712948,
|
| 1763 |
+
"grad_norm": 2.037355422973633,
|
| 1764 |
+
"learning_rate": 9.183037464140804e-07,
|
| 1765 |
+
"loss": 0.3457,
|
| 1766 |
+
"step": 2510
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 7.85179407176287,
|
| 1770 |
+
"grad_norm": 1.6725716590881348,
|
| 1771 |
+
"learning_rate": 8.823370118578628e-07,
|
| 1772 |
+
"loss": 0.3437,
|
| 1773 |
+
"step": 2520
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 7.8829953198127924,
|
| 1777 |
+
"grad_norm": 1.7678967714309692,
|
| 1778 |
+
"learning_rate": 8.470205040209362e-07,
|
| 1779 |
+
"loss": 0.3022,
|
| 1780 |
+
"step": 2530
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 7.914196567862715,
|
| 1784 |
+
"grad_norm": 2.164024591445923,
|
| 1785 |
+
"learning_rate": 8.123597997892918e-07,
|
| 1786 |
+
"loss": 0.3465,
|
| 1787 |
+
"step": 2540
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 7.945397815912637,
|
| 1791 |
+
"grad_norm": 1.7257033586502075,
|
| 1792 |
+
"learning_rate": 7.783603724899258e-07,
|
| 1793 |
+
"loss": 0.3393,
|
| 1794 |
+
"step": 2550
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 7.976599063962558,
|
| 1798 |
+
"grad_norm": 1.8956247568130493,
|
| 1799 |
+
"learning_rate": 7.450275910265415e-07,
|
| 1800 |
+
"loss": 0.3502,
|
| 1801 |
+
"step": 2560
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 8.006240249609984,
|
| 1805 |
+
"grad_norm": 1.7082842588424683,
|
| 1806 |
+
"learning_rate": 7.123667190317396e-07,
|
| 1807 |
+
"loss": 0.3473,
|
| 1808 |
+
"step": 2570
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 8.037441497659906,
|
| 1812 |
+
"grad_norm": 1.825515866279602,
|
| 1813 |
+
"learning_rate": 6.803829140358237e-07,
|
| 1814 |
+
"loss": 0.3615,
|
| 1815 |
+
"step": 2580
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 8.068642745709829,
|
| 1819 |
+
"grad_norm": 1.7333165407180786,
|
| 1820 |
+
"learning_rate": 6.490812266523716e-07,
|
| 1821 |
+
"loss": 0.355,
|
| 1822 |
+
"step": 2590
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 8.09984399375975,
|
| 1826 |
+
"grad_norm": 1.7681299448013306,
|
| 1827 |
+
"learning_rate": 6.184665997806827e-07,
|
| 1828 |
+
"loss": 0.3628,
|
| 1829 |
+
"step": 2600
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 8.131045241809673,
|
| 1833 |
+
"grad_norm": 1.8737382888793945,
|
| 1834 |
+
"learning_rate": 5.885438678252342e-07,
|
| 1835 |
+
"loss": 0.367,
|
| 1836 |
+
"step": 2610
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 8.162246489859594,
|
| 1840 |
+
"grad_norm": 1.9339786767959595,
|
| 1841 |
+
"learning_rate": 5.593177559322776e-07,
|
| 1842 |
+
"loss": 0.3652,
|
| 1843 |
+
"step": 2620
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 8.193447737909516,
|
| 1847 |
+
"grad_norm": 1.9356807470321655,
|
| 1848 |
+
"learning_rate": 5.307928792436812e-07,
|
| 1849 |
+
"loss": 0.3431,
|
| 1850 |
+
"step": 2630
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 8.224648985959439,
|
| 1854 |
+
"grad_norm": 1.654337763786316,
|
| 1855 |
+
"learning_rate": 5.029737421681446e-07,
|
| 1856 |
+
"loss": 0.3467,
|
| 1857 |
+
"step": 2640
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 8.25585023400936,
|
| 1861 |
+
"grad_norm": 1.8306303024291992,
|
| 1862 |
+
"learning_rate": 4.758647376699033e-07,
|
| 1863 |
+
"loss": 0.3191,
|
| 1864 |
+
"step": 2650
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 8.287051482059283,
|
| 1868 |
+
"grad_norm": 1.7232128381729126,
|
| 1869 |
+
"learning_rate": 4.494701465750217e-07,
|
| 1870 |
+
"loss": 0.3412,
|
| 1871 |
+
"step": 2660
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 8.318252730109204,
|
| 1875 |
+
"grad_norm": 1.7647130489349365,
|
| 1876 |
+
"learning_rate": 4.237941368954124e-07,
|
| 1877 |
+
"loss": 0.317,
|
| 1878 |
+
"step": 2670
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 8.349453978159126,
|
| 1882 |
+
"grad_norm": 1.6271697282791138,
|
| 1883 |
+
"learning_rate": 3.9884076317064813e-07,
|
| 1884 |
+
"loss": 0.3361,
|
| 1885 |
+
"step": 2680
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 8.380655226209049,
|
| 1889 |
+
"grad_norm": 1.615136981010437,
|
| 1890 |
+
"learning_rate": 3.7461396582771035e-07,
|
| 1891 |
+
"loss": 0.3417,
|
| 1892 |
+
"step": 2690
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 8.41185647425897,
|
| 1896 |
+
"grad_norm": 2.1233370304107666,
|
| 1897 |
+
"learning_rate": 3.511175705587433e-07,
|
| 1898 |
+
"loss": 0.3601,
|
| 1899 |
+
"step": 2700
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 8.443057722308893,
|
| 1903 |
+
"grad_norm": 1.7187414169311523,
|
| 1904 |
+
"learning_rate": 3.283552877169399e-07,
|
| 1905 |
+
"loss": 0.3739,
|
| 1906 |
+
"step": 2710
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 8.474258970358814,
|
| 1910 |
+
"grad_norm": 1.9059900045394897,
|
| 1911 |
+
"learning_rate": 3.0633071173062966e-07,
|
| 1912 |
+
"loss": 0.3205,
|
| 1913 |
+
"step": 2720
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 8.505460218408736,
|
| 1917 |
+
"grad_norm": 1.6844722032546997,
|
| 1918 |
+
"learning_rate": 2.850473205356774e-07,
|
| 1919 |
+
"loss": 0.3565,
|
| 1920 |
+
"step": 2730
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 8.536661466458659,
|
| 1924 |
+
"grad_norm": 1.8269972801208496,
|
| 1925 |
+
"learning_rate": 2.6450847502627883e-07,
|
| 1926 |
+
"loss": 0.3435,
|
| 1927 |
+
"step": 2740
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 8.56786271450858,
|
| 1931 |
+
"grad_norm": 1.8859301805496216,
|
| 1932 |
+
"learning_rate": 2.447174185242324e-07,
|
| 1933 |
+
"loss": 0.3781,
|
| 1934 |
+
"step": 2750
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 8.599063962558503,
|
| 1938 |
+
"grad_norm": 1.8474805355072021,
|
| 1939 |
+
"learning_rate": 2.2567727626678527e-07,
|
| 1940 |
+
"loss": 0.3235,
|
| 1941 |
+
"step": 2760
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 8.630265210608425,
|
| 1945 |
+
"grad_norm": 1.8365707397460938,
|
| 1946 |
+
"learning_rate": 2.0739105491312028e-07,
|
| 1947 |
+
"loss": 0.3484,
|
| 1948 |
+
"step": 2770
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 8.661466458658346,
|
| 1952 |
+
"grad_norm": 1.6072348356246948,
|
| 1953 |
+
"learning_rate": 1.8986164206957037e-07,
|
| 1954 |
+
"loss": 0.3482,
|
| 1955 |
+
"step": 2780
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 8.692667706708269,
|
| 1959 |
+
"grad_norm": 1.8976094722747803,
|
| 1960 |
+
"learning_rate": 1.7309180583363062e-07,
|
| 1961 |
+
"loss": 0.3423,
|
| 1962 |
+
"step": 2790
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 8.72386895475819,
|
| 1966 |
+
"grad_norm": 1.7271218299865723,
|
| 1967 |
+
"learning_rate": 1.5708419435684463e-07,
|
| 1968 |
+
"loss": 0.3661,
|
| 1969 |
+
"step": 2800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 8.755070202808112,
|
| 1973 |
+
"grad_norm": 1.6312350034713745,
|
| 1974 |
+
"learning_rate": 1.4184133542663014e-07,
|
| 1975 |
+
"loss": 0.3577,
|
| 1976 |
+
"step": 2810
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 8.786271450858035,
|
| 1980 |
+
"grad_norm": 1.7652922868728638,
|
| 1981 |
+
"learning_rate": 1.2736563606711384e-07,
|
| 1982 |
+
"loss": 0.3469,
|
| 1983 |
+
"step": 2820
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 8.817472698907956,
|
| 1987 |
+
"grad_norm": 1.7783547639846802,
|
| 1988 |
+
"learning_rate": 1.136593821590326e-07,
|
| 1989 |
+
"loss": 0.365,
|
| 1990 |
+
"step": 2830
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 8.848673946957879,
|
| 1994 |
+
"grad_norm": 1.8157516717910767,
|
| 1995 |
+
"learning_rate": 1.007247380787657e-07,
|
| 1996 |
+
"loss": 0.3469,
|
| 1997 |
+
"step": 2840
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 8.8798751950078,
|
| 2001 |
+
"grad_norm": 1.8091756105422974,
|
| 2002 |
+
"learning_rate": 8.856374635655696e-08,
|
| 2003 |
+
"loss": 0.3255,
|
| 2004 |
+
"step": 2850
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 8.911076443057722,
|
| 2008 |
+
"grad_norm": 1.5107375383377075,
|
| 2009 |
+
"learning_rate": 7.717832735397335e-08,
|
| 2010 |
+
"loss": 0.3576,
|
| 2011 |
+
"step": 2860
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 8.942277691107645,
|
| 2015 |
+
"grad_norm": 1.7504184246063232,
|
| 2016 |
+
"learning_rate": 6.657027896065982e-08,
|
| 2017 |
+
"loss": 0.3657,
|
| 2018 |
+
"step": 2870
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 8.973478939157566,
|
| 2022 |
+
"grad_norm": 1.8140274286270142,
|
| 2023 |
+
"learning_rate": 5.674127631043025e-08,
|
| 2024 |
+
"loss": 0.3513,
|
| 2025 |
+
"step": 2880
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 9.003120124804992,
|
| 2029 |
+
"grad_norm": 1.6057106256484985,
|
| 2030 |
+
"learning_rate": 4.769287151674407e-08,
|
| 2031 |
+
"loss": 0.336,
|
| 2032 |
+
"step": 2890
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 9.034321372854913,
|
| 2036 |
+
"grad_norm": 1.670082926750183,
|
| 2037 |
+
"learning_rate": 3.9426493427611177e-08,
|
| 2038 |
+
"loss": 0.3572,
|
| 2039 |
+
"step": 2900
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 9.065522620904837,
|
| 2043 |
+
"grad_norm": 1.5580389499664307,
|
| 2044 |
+
"learning_rate": 3.194344739995803e-08,
|
| 2045 |
+
"loss": 0.3224,
|
| 2046 |
+
"step": 2910
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 9.096723868954758,
|
| 2050 |
+
"grad_norm": 2.9034836292266846,
|
| 2051 |
+
"learning_rate": 2.5244915093499134e-08,
|
| 2052 |
+
"loss": 0.3334,
|
| 2053 |
+
"step": 2920
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 9.127925117004681,
|
| 2057 |
+
"grad_norm": 1.8099640607833862,
|
| 2058 |
+
"learning_rate": 1.9331954284137476e-08,
|
| 2059 |
+
"loss": 0.3383,
|
| 2060 |
+
"step": 2930
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 9.159126365054602,
|
| 2064 |
+
"grad_norm": 1.7415984869003296,
|
| 2065 |
+
"learning_rate": 1.4205498696930332e-08,
|
| 2066 |
+
"loss": 0.3366,
|
| 2067 |
+
"step": 2940
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 9.190327613104524,
|
| 2071 |
+
"grad_norm": 1.813762903213501,
|
| 2072 |
+
"learning_rate": 9.866357858642206e-09,
|
| 2073 |
+
"loss": 0.3436,
|
| 2074 |
+
"step": 2950
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 9.221528861154447,
|
| 2078 |
+
"grad_norm": 1.8357665538787842,
|
| 2079 |
+
"learning_rate": 6.315216969912663e-09,
|
| 2080 |
+
"loss": 0.3617,
|
| 2081 |
+
"step": 2960
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 9.252730109204368,
|
| 2085 |
+
"grad_norm": 2.055318593978882,
|
| 2086 |
+
"learning_rate": 3.5526367970539765e-09,
|
| 2087 |
+
"loss": 0.3564,
|
| 2088 |
+
"step": 2970
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 9.283931357254291,
|
| 2092 |
+
"grad_norm": 1.8779317140579224,
|
| 2093 |
+
"learning_rate": 1.5790535835003006e-09,
|
| 2094 |
+
"loss": 0.3498,
|
| 2095 |
+
"step": 2980
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 9.315132605304212,
|
| 2099 |
+
"grad_norm": 1.920357584953308,
|
| 2100 |
+
"learning_rate": 3.9477898091944135e-10,
|
| 2101 |
+
"loss": 0.3099,
|
| 2102 |
+
"step": 2990
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 9.346333853354134,
|
| 2106 |
+
"grad_norm": 1.7496743202209473,
|
| 2107 |
+
"learning_rate": 0.0,
|
| 2108 |
+
"loss": 0.3488,
|
| 2109 |
+
"step": 3000
|
| 2110 |
+
}
|
| 2111 |
+
],
|
| 2112 |
+
"logging_steps": 10,
|
| 2113 |
+
"max_steps": 3000,
|
| 2114 |
+
"num_input_tokens_seen": 0,
|
| 2115 |
+
"num_train_epochs": 10,
|
| 2116 |
+
"save_steps": 1000,
|
| 2117 |
+
"stateful_callbacks": {
|
| 2118 |
+
"TrainerControl": {
|
| 2119 |
+
"args": {
|
| 2120 |
+
"should_epoch_stop": false,
|
| 2121 |
+
"should_evaluate": false,
|
| 2122 |
+
"should_log": false,
|
| 2123 |
+
"should_save": true,
|
| 2124 |
+
"should_training_stop": true
|
| 2125 |
+
},
|
| 2126 |
+
"attributes": {}
|
| 2127 |
+
}
|
| 2128 |
+
},
|
| 2129 |
+
"total_flos": 8.902153377467597e+16,
|
| 2130 |
+
"train_batch_size": 4,
|
| 2131 |
+
"trial_name": null,
|
| 2132 |
+
"trial_params": null
|
| 2133 |
+
}
|
amt_PKMN_Harmonizer_Medium/checkpoint-3000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:620747907b2130f2f54d285e74e3eb145407d8bba5652ffcf90771ba77206c41
|
| 3 |
+
size 5713
|