init v0.1.0
Browse files- config.json +29 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +760 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation": "gelu",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DistilBertForSequenceClassification"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"dim": 768,
|
| 8 |
+
"dropout": 0.1,
|
| 9 |
+
"hidden_dim": 3072,
|
| 10 |
+
"id2label": {
|
| 11 |
+
"0": "LABEL_0"
|
| 12 |
+
},
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"label2id": {
|
| 15 |
+
"LABEL_0": 0
|
| 16 |
+
},
|
| 17 |
+
"max_position_embeddings": 512,
|
| 18 |
+
"model_type": "distilbert",
|
| 19 |
+
"n_heads": 12,
|
| 20 |
+
"n_layers": 6,
|
| 21 |
+
"pad_token_id": 0,
|
| 22 |
+
"qa_dropout": 0.1,
|
| 23 |
+
"seq_classif_dropout": 0.2,
|
| 24 |
+
"sinusoidal_pos_embds": true,
|
| 25 |
+
"tie_weights_": true,
|
| 26 |
+
"torch_dtype": "float32",
|
| 27 |
+
"transformers_version": "4.53.2",
|
| 28 |
+
"vocab_size": 32000
|
| 29 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:002eebd87f40fc8936783f9e427a1dccd7e7e60d762ca8721703299aedd71070
|
| 3 |
+
size 272369900
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ef68dff28032e5acf3716fdc317320a4c4c0fb1ec9227a6de40216cd9e0d37d
|
| 3 |
+
size 544799563
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:791590502d32babfec3e01cad84acac1a5c5f69449f6851db53f4aead2041f79
|
| 3 |
+
size 14455
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88e67103b7bacb3746e8bd09a33738fe2674f4ced3f3c7714e2384f23492cdd4
|
| 3 |
+
size 1465
|
trainer_state.json
ADDED
|
@@ -0,0 +1,760 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1326,
|
| 3 |
+
"best_metric": 1.3003416061401367,
|
| 4 |
+
"best_model_checkpoint": "./results/checkpoint-1326",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4420,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.11312217194570136,
|
| 14 |
+
"grad_norm": 1.7515817880630493,
|
| 15 |
+
"learning_rate": 9.889140271493214e-06,
|
| 16 |
+
"loss": 1.9771,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.22624434389140272,
|
| 21 |
+
"grad_norm": 0.3798021972179413,
|
| 22 |
+
"learning_rate": 9.776018099547512e-06,
|
| 23 |
+
"loss": 3.0565,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.3393665158371041,
|
| 28 |
+
"grad_norm": 0.24794337153434753,
|
| 29 |
+
"learning_rate": 9.662895927601811e-06,
|
| 30 |
+
"loss": 3.1013,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.45248868778280543,
|
| 35 |
+
"grad_norm": 0.13749714195728302,
|
| 36 |
+
"learning_rate": 9.54977375565611e-06,
|
| 37 |
+
"loss": 2.8378,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.5656108597285068,
|
| 42 |
+
"grad_norm": 0.1238800510764122,
|
| 43 |
+
"learning_rate": 9.436651583710407e-06,
|
| 44 |
+
"loss": 2.1715,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.6787330316742082,
|
| 49 |
+
"grad_norm": 0.24898090958595276,
|
| 50 |
+
"learning_rate": 9.323529411764707e-06,
|
| 51 |
+
"loss": 3.7122,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.7918552036199095,
|
| 56 |
+
"grad_norm": 0.23299390077590942,
|
| 57 |
+
"learning_rate": 9.210407239819005e-06,
|
| 58 |
+
"loss": 2.6715,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.9049773755656109,
|
| 63 |
+
"grad_norm": 0.07253708690404892,
|
| 64 |
+
"learning_rate": 9.097285067873303e-06,
|
| 65 |
+
"loss": 1.4325,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.0,
|
| 70 |
+
"eval_f1": 0.5660377358490566,
|
| 71 |
+
"eval_loss": 2.0750715732574463,
|
| 72 |
+
"eval_precision": 0.6818181818181818,
|
| 73 |
+
"eval_recall": 0.4838709677419355,
|
| 74 |
+
"eval_runtime": 8.6034,
|
| 75 |
+
"eval_samples_per_second": 205.499,
|
| 76 |
+
"eval_steps_per_second": 12.902,
|
| 77 |
+
"step": 442
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.0180995475113122,
|
| 81 |
+
"grad_norm": 223.5320281982422,
|
| 82 |
+
"learning_rate": 8.984162895927603e-06,
|
| 83 |
+
"loss": 2.09,
|
| 84 |
+
"step": 450
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.1312217194570136,
|
| 88 |
+
"grad_norm": 0.810788094997406,
|
| 89 |
+
"learning_rate": 8.8710407239819e-06,
|
| 90 |
+
"loss": 1.6058,
|
| 91 |
+
"step": 500
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.244343891402715,
|
| 95 |
+
"grad_norm": 0.12611542642116547,
|
| 96 |
+
"learning_rate": 8.7579185520362e-06,
|
| 97 |
+
"loss": 1.116,
|
| 98 |
+
"step": 550
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.3574660633484164,
|
| 102 |
+
"grad_norm": 0.0376112200319767,
|
| 103 |
+
"learning_rate": 8.644796380090498e-06,
|
| 104 |
+
"loss": 0.5762,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.4705882352941178,
|
| 109 |
+
"grad_norm": 0.060051899403333664,
|
| 110 |
+
"learning_rate": 8.531674208144796e-06,
|
| 111 |
+
"loss": 2.5467,
|
| 112 |
+
"step": 650
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.5837104072398192,
|
| 116 |
+
"grad_norm": 0.07170303165912628,
|
| 117 |
+
"learning_rate": 8.418552036199096e-06,
|
| 118 |
+
"loss": 1.1257,
|
| 119 |
+
"step": 700
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 1.6968325791855203,
|
| 123 |
+
"grad_norm": 0.4100867211818695,
|
| 124 |
+
"learning_rate": 8.305429864253394e-06,
|
| 125 |
+
"loss": 2.1582,
|
| 126 |
+
"step": 750
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 1.8099547511312217,
|
| 130 |
+
"grad_norm": 0.034761942923069,
|
| 131 |
+
"learning_rate": 8.192307692307692e-06,
|
| 132 |
+
"loss": 1.2032,
|
| 133 |
+
"step": 800
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.9230769230769231,
|
| 137 |
+
"grad_norm": 0.07466017454862595,
|
| 138 |
+
"learning_rate": 8.079185520361992e-06,
|
| 139 |
+
"loss": 0.85,
|
| 140 |
+
"step": 850
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.0,
|
| 144 |
+
"eval_f1": 0.5357142857142857,
|
| 145 |
+
"eval_loss": 1.7785077095031738,
|
| 146 |
+
"eval_precision": 0.6,
|
| 147 |
+
"eval_recall": 0.4838709677419355,
|
| 148 |
+
"eval_runtime": 8.5452,
|
| 149 |
+
"eval_samples_per_second": 206.9,
|
| 150 |
+
"eval_steps_per_second": 12.99,
|
| 151 |
+
"step": 884
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.0361990950226243,
|
| 155 |
+
"grad_norm": 0.07509780675172806,
|
| 156 |
+
"learning_rate": 7.96606334841629e-06,
|
| 157 |
+
"loss": 1.8895,
|
| 158 |
+
"step": 900
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.1493212669683257,
|
| 162 |
+
"grad_norm": 0.062317393720149994,
|
| 163 |
+
"learning_rate": 7.85294117647059e-06,
|
| 164 |
+
"loss": 0.5769,
|
| 165 |
+
"step": 950
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.262443438914027,
|
| 169 |
+
"grad_norm": 0.26913294196128845,
|
| 170 |
+
"learning_rate": 7.739819004524888e-06,
|
| 171 |
+
"loss": 1.0833,
|
| 172 |
+
"step": 1000
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.3755656108597285,
|
| 176 |
+
"grad_norm": 4.26383638381958,
|
| 177 |
+
"learning_rate": 7.626696832579186e-06,
|
| 178 |
+
"loss": 0.6466,
|
| 179 |
+
"step": 1050
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.48868778280543,
|
| 183 |
+
"grad_norm": 5.8240509033203125,
|
| 184 |
+
"learning_rate": 7.5135746606334844e-06,
|
| 185 |
+
"loss": 1.3183,
|
| 186 |
+
"step": 1100
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.6018099547511313,
|
| 190 |
+
"grad_norm": 1.614251732826233,
|
| 191 |
+
"learning_rate": 7.400452488687784e-06,
|
| 192 |
+
"loss": 0.6002,
|
| 193 |
+
"step": 1150
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.7149321266968327,
|
| 197 |
+
"grad_norm": 0.018574368208646774,
|
| 198 |
+
"learning_rate": 7.287330316742081e-06,
|
| 199 |
+
"loss": 1.1073,
|
| 200 |
+
"step": 1200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 2.8280542986425337,
|
| 204 |
+
"grad_norm": 1.8890864849090576,
|
| 205 |
+
"learning_rate": 7.174208144796381e-06,
|
| 206 |
+
"loss": 0.9746,
|
| 207 |
+
"step": 1250
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 2.9411764705882355,
|
| 211 |
+
"grad_norm": 1.8844258785247803,
|
| 212 |
+
"learning_rate": 7.06108597285068e-06,
|
| 213 |
+
"loss": 1.1957,
|
| 214 |
+
"step": 1300
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 3.0,
|
| 218 |
+
"eval_f1": 0.6857142857142857,
|
| 219 |
+
"eval_loss": 1.3003416061401367,
|
| 220 |
+
"eval_precision": 0.6153846153846154,
|
| 221 |
+
"eval_recall": 0.7741935483870968,
|
| 222 |
+
"eval_runtime": 8.5419,
|
| 223 |
+
"eval_samples_per_second": 206.98,
|
| 224 |
+
"eval_steps_per_second": 12.995,
|
| 225 |
+
"step": 1326
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.0542986425339365,
|
| 229 |
+
"grad_norm": 0.02057269588112831,
|
| 230 |
+
"learning_rate": 6.947963800904978e-06,
|
| 231 |
+
"loss": 0.9242,
|
| 232 |
+
"step": 1350
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.167420814479638,
|
| 236 |
+
"grad_norm": 0.016314402222633362,
|
| 237 |
+
"learning_rate": 6.834841628959277e-06,
|
| 238 |
+
"loss": 1.1972,
|
| 239 |
+
"step": 1400
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.2805429864253393,
|
| 243 |
+
"grad_norm": 0.016238074749708176,
|
| 244 |
+
"learning_rate": 6.7217194570135755e-06,
|
| 245 |
+
"loss": 0.5929,
|
| 246 |
+
"step": 1450
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 3.3936651583710407,
|
| 250 |
+
"grad_norm": 2.082648992538452,
|
| 251 |
+
"learning_rate": 6.6085972850678735e-06,
|
| 252 |
+
"loss": 0.0265,
|
| 253 |
+
"step": 1500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 3.506787330316742,
|
| 257 |
+
"grad_norm": 17.545307159423828,
|
| 258 |
+
"learning_rate": 6.495475113122172e-06,
|
| 259 |
+
"loss": 1.7171,
|
| 260 |
+
"step": 1550
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 3.6199095022624435,
|
| 264 |
+
"grad_norm": 0.6719046831130981,
|
| 265 |
+
"learning_rate": 6.38235294117647e-06,
|
| 266 |
+
"loss": 0.0379,
|
| 267 |
+
"step": 1600
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.733031674208145,
|
| 271 |
+
"grad_norm": 3.1016860008239746,
|
| 272 |
+
"learning_rate": 6.26923076923077e-06,
|
| 273 |
+
"loss": 0.7337,
|
| 274 |
+
"step": 1650
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 3.8461538461538463,
|
| 278 |
+
"grad_norm": 0.12171656638383865,
|
| 279 |
+
"learning_rate": 6.156108597285069e-06,
|
| 280 |
+
"loss": 0.0274,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 3.9592760180995477,
|
| 285 |
+
"grad_norm": 3.5538485050201416,
|
| 286 |
+
"learning_rate": 6.042986425339367e-06,
|
| 287 |
+
"loss": 0.7957,
|
| 288 |
+
"step": 1750
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 4.0,
|
| 292 |
+
"eval_f1": 0.6956521739130435,
|
| 293 |
+
"eval_loss": 1.482049822807312,
|
| 294 |
+
"eval_precision": 0.631578947368421,
|
| 295 |
+
"eval_recall": 0.7741935483870968,
|
| 296 |
+
"eval_runtime": 8.6318,
|
| 297 |
+
"eval_samples_per_second": 204.824,
|
| 298 |
+
"eval_steps_per_second": 12.859,
|
| 299 |
+
"step": 1768
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.072398190045249,
|
| 303 |
+
"grad_norm": 0.01248278096318245,
|
| 304 |
+
"learning_rate": 5.929864253393666e-06,
|
| 305 |
+
"loss": 0.7697,
|
| 306 |
+
"step": 1800
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.1855203619909505,
|
| 310 |
+
"grad_norm": 0.026634668931365013,
|
| 311 |
+
"learning_rate": 5.816742081447965e-06,
|
| 312 |
+
"loss": 0.9681,
|
| 313 |
+
"step": 1850
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 4.298642533936651,
|
| 317 |
+
"grad_norm": 0.011785144917666912,
|
| 318 |
+
"learning_rate": 5.703619909502263e-06,
|
| 319 |
+
"loss": 0.0388,
|
| 320 |
+
"step": 1900
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 4.411764705882353,
|
| 324 |
+
"grad_norm": 0.07407853752374649,
|
| 325 |
+
"learning_rate": 5.5904977375565615e-06,
|
| 326 |
+
"loss": 0.549,
|
| 327 |
+
"step": 1950
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 4.524886877828054,
|
| 331 |
+
"grad_norm": 0.4233396649360657,
|
| 332 |
+
"learning_rate": 5.47737556561086e-06,
|
| 333 |
+
"loss": 0.4991,
|
| 334 |
+
"step": 2000
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 4.638009049773755,
|
| 338 |
+
"grad_norm": 0.00987264234572649,
|
| 339 |
+
"learning_rate": 5.364253393665158e-06,
|
| 340 |
+
"loss": 0.0074,
|
| 341 |
+
"step": 2050
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 4.751131221719457,
|
| 345 |
+
"grad_norm": 0.08375240862369537,
|
| 346 |
+
"learning_rate": 5.251131221719458e-06,
|
| 347 |
+
"loss": 0.9549,
|
| 348 |
+
"step": 2100
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 4.864253393665159,
|
| 352 |
+
"grad_norm": 1.7564586400985718,
|
| 353 |
+
"learning_rate": 5.138009049773756e-06,
|
| 354 |
+
"loss": 0.5174,
|
| 355 |
+
"step": 2150
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 4.97737556561086,
|
| 359 |
+
"grad_norm": 0.21808204054832458,
|
| 360 |
+
"learning_rate": 5.024886877828055e-06,
|
| 361 |
+
"loss": 0.6401,
|
| 362 |
+
"step": 2200
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 5.0,
|
| 366 |
+
"eval_f1": 0.6666666666666666,
|
| 367 |
+
"eval_loss": 1.3885061740875244,
|
| 368 |
+
"eval_precision": 0.5853658536585366,
|
| 369 |
+
"eval_recall": 0.7741935483870968,
|
| 370 |
+
"eval_runtime": 8.8942,
|
| 371 |
+
"eval_samples_per_second": 198.781,
|
| 372 |
+
"eval_steps_per_second": 12.48,
|
| 373 |
+
"step": 2210
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.090497737556561,
|
| 377 |
+
"grad_norm": 0.009224635548889637,
|
| 378 |
+
"learning_rate": 4.911764705882353e-06,
|
| 379 |
+
"loss": 0.2408,
|
| 380 |
+
"step": 2250
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.203619909502263,
|
| 384 |
+
"grad_norm": 0.14075350761413574,
|
| 385 |
+
"learning_rate": 4.7986425339366525e-06,
|
| 386 |
+
"loss": 0.1984,
|
| 387 |
+
"step": 2300
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 5.316742081447964,
|
| 391 |
+
"grad_norm": 0.011622537858784199,
|
| 392 |
+
"learning_rate": 4.6855203619909505e-06,
|
| 393 |
+
"loss": 0.334,
|
| 394 |
+
"step": 2350
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 5.429864253393665,
|
| 398 |
+
"grad_norm": 166.04515075683594,
|
| 399 |
+
"learning_rate": 4.572398190045249e-06,
|
| 400 |
+
"loss": 0.5887,
|
| 401 |
+
"step": 2400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 5.542986425339366,
|
| 405 |
+
"grad_norm": 0.0077699883840978146,
|
| 406 |
+
"learning_rate": 4.459276018099548e-06,
|
| 407 |
+
"loss": 0.6319,
|
| 408 |
+
"step": 2450
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 5.656108597285068,
|
| 412 |
+
"grad_norm": 0.3266526758670807,
|
| 413 |
+
"learning_rate": 4.346153846153846e-06,
|
| 414 |
+
"loss": 0.0257,
|
| 415 |
+
"step": 2500
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 5.769230769230769,
|
| 419 |
+
"grad_norm": 0.007749281823635101,
|
| 420 |
+
"learning_rate": 4.233031674208145e-06,
|
| 421 |
+
"loss": 0.1457,
|
| 422 |
+
"step": 2550
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 5.882352941176471,
|
| 426 |
+
"grad_norm": 0.006045708432793617,
|
| 427 |
+
"learning_rate": 4.119909502262444e-06,
|
| 428 |
+
"loss": 0.3572,
|
| 429 |
+
"step": 2600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 5.995475113122172,
|
| 433 |
+
"grad_norm": 0.010532204993069172,
|
| 434 |
+
"learning_rate": 4.006787330316743e-06,
|
| 435 |
+
"loss": 1.0198,
|
| 436 |
+
"step": 2650
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 6.0,
|
| 440 |
+
"eval_f1": 0.631578947368421,
|
| 441 |
+
"eval_loss": 1.4059184789657593,
|
| 442 |
+
"eval_precision": 0.5333333333333333,
|
| 443 |
+
"eval_recall": 0.7741935483870968,
|
| 444 |
+
"eval_runtime": 8.5447,
|
| 445 |
+
"eval_samples_per_second": 206.912,
|
| 446 |
+
"eval_steps_per_second": 12.991,
|
| 447 |
+
"step": 2652
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 6.108597285067873,
|
| 451 |
+
"grad_norm": 0.010519157163798809,
|
| 452 |
+
"learning_rate": 3.893665158371041e-06,
|
| 453 |
+
"loss": 0.0923,
|
| 454 |
+
"step": 2700
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 6.221719457013575,
|
| 458 |
+
"grad_norm": 0.010456902906298637,
|
| 459 |
+
"learning_rate": 3.7805429864253396e-06,
|
| 460 |
+
"loss": 0.0487,
|
| 461 |
+
"step": 2750
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 6.334841628959276,
|
| 465 |
+
"grad_norm": 0.18975257873535156,
|
| 466 |
+
"learning_rate": 3.667420814479638e-06,
|
| 467 |
+
"loss": 0.0116,
|
| 468 |
+
"step": 2800
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 6.447963800904978,
|
| 472 |
+
"grad_norm": 25.184865951538086,
|
| 473 |
+
"learning_rate": 3.554298642533937e-06,
|
| 474 |
+
"loss": 0.1499,
|
| 475 |
+
"step": 2850
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 6.5610859728506785,
|
| 479 |
+
"grad_norm": 1.7686313390731812,
|
| 480 |
+
"learning_rate": 3.4411764705882358e-06,
|
| 481 |
+
"loss": 0.5646,
|
| 482 |
+
"step": 2900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 6.67420814479638,
|
| 486 |
+
"grad_norm": 0.005733998026698828,
|
| 487 |
+
"learning_rate": 3.328054298642534e-06,
|
| 488 |
+
"loss": 0.4556,
|
| 489 |
+
"step": 2950
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 6.787330316742081,
|
| 493 |
+
"grad_norm": 2.235478639602661,
|
| 494 |
+
"learning_rate": 3.214932126696833e-06,
|
| 495 |
+
"loss": 0.0495,
|
| 496 |
+
"step": 3000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 6.900452488687783,
|
| 500 |
+
"grad_norm": 0.014012620784342289,
|
| 501 |
+
"learning_rate": 3.1018099547511315e-06,
|
| 502 |
+
"loss": 0.9254,
|
| 503 |
+
"step": 3050
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 7.0,
|
| 507 |
+
"eval_f1": 0.631578947368421,
|
| 508 |
+
"eval_loss": 1.4806020259857178,
|
| 509 |
+
"eval_precision": 0.5333333333333333,
|
| 510 |
+
"eval_recall": 0.7741935483870968,
|
| 511 |
+
"eval_runtime": 8.5485,
|
| 512 |
+
"eval_samples_per_second": 206.819,
|
| 513 |
+
"eval_steps_per_second": 12.985,
|
| 514 |
+
"step": 3094
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 7.013574660633484,
|
| 518 |
+
"grad_norm": 0.14705590903759003,
|
| 519 |
+
"learning_rate": 2.98868778280543e-06,
|
| 520 |
+
"loss": 0.2657,
|
| 521 |
+
"step": 3100
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 7.126696832579185,
|
| 525 |
+
"grad_norm": 0.01319128554314375,
|
| 526 |
+
"learning_rate": 2.8755656108597287e-06,
|
| 527 |
+
"loss": 0.0663,
|
| 528 |
+
"step": 3150
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 7.239819004524887,
|
| 532 |
+
"grad_norm": 0.006507423706352711,
|
| 533 |
+
"learning_rate": 2.7624434389140276e-06,
|
| 534 |
+
"loss": 0.4037,
|
| 535 |
+
"step": 3200
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 7.352941176470588,
|
| 539 |
+
"grad_norm": 0.006867765448987484,
|
| 540 |
+
"learning_rate": 2.649321266968326e-06,
|
| 541 |
+
"loss": 0.4015,
|
| 542 |
+
"step": 3250
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 7.46606334841629,
|
| 546 |
+
"grad_norm": 0.09424237906932831,
|
| 547 |
+
"learning_rate": 2.5361990950226244e-06,
|
| 548 |
+
"loss": 0.5639,
|
| 549 |
+
"step": 3300
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 7.579185520361991,
|
| 553 |
+
"grad_norm": 0.1817472279071808,
|
| 554 |
+
"learning_rate": 2.4230769230769233e-06,
|
| 555 |
+
"loss": 0.0205,
|
| 556 |
+
"step": 3350
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 7.6923076923076925,
|
| 560 |
+
"grad_norm": 0.006896049249917269,
|
| 561 |
+
"learning_rate": 2.309954751131222e-06,
|
| 562 |
+
"loss": 0.0364,
|
| 563 |
+
"step": 3400
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 7.8054298642533935,
|
| 567 |
+
"grad_norm": 0.005642372649163008,
|
| 568 |
+
"learning_rate": 2.1968325791855205e-06,
|
| 569 |
+
"loss": 0.5472,
|
| 570 |
+
"step": 3450
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 7.918552036199095,
|
| 574 |
+
"grad_norm": 0.21695345640182495,
|
| 575 |
+
"learning_rate": 2.0837104072398194e-06,
|
| 576 |
+
"loss": 0.3692,
|
| 577 |
+
"step": 3500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 8.0,
|
| 581 |
+
"eval_f1": 0.676056338028169,
|
| 582 |
+
"eval_loss": 1.7710559368133545,
|
| 583 |
+
"eval_precision": 0.6,
|
| 584 |
+
"eval_recall": 0.7741935483870968,
|
| 585 |
+
"eval_runtime": 8.516,
|
| 586 |
+
"eval_samples_per_second": 207.61,
|
| 587 |
+
"eval_steps_per_second": 13.034,
|
| 588 |
+
"step": 3536
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 8.031674208144796,
|
| 592 |
+
"grad_norm": 0.004109715577214956,
|
| 593 |
+
"learning_rate": 1.970588235294118e-06,
|
| 594 |
+
"loss": 0.0172,
|
| 595 |
+
"step": 3550
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 8.144796380090497,
|
| 599 |
+
"grad_norm": 0.07877568900585175,
|
| 600 |
+
"learning_rate": 1.8574660633484164e-06,
|
| 601 |
+
"loss": 0.0129,
|
| 602 |
+
"step": 3600
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 8.2579185520362,
|
| 606 |
+
"grad_norm": 0.004373373929411173,
|
| 607 |
+
"learning_rate": 1.744343891402715e-06,
|
| 608 |
+
"loss": 0.0199,
|
| 609 |
+
"step": 3650
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 8.371040723981901,
|
| 613 |
+
"grad_norm": 0.004394580144435167,
|
| 614 |
+
"learning_rate": 1.6312217194570137e-06,
|
| 615 |
+
"loss": 0.287,
|
| 616 |
+
"step": 3700
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 8.484162895927602,
|
| 620 |
+
"grad_norm": 2.0952539443969727,
|
| 621 |
+
"learning_rate": 1.5180995475113121e-06,
|
| 622 |
+
"loss": 0.0485,
|
| 623 |
+
"step": 3750
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 8.597285067873303,
|
| 627 |
+
"grad_norm": 1487.0665283203125,
|
| 628 |
+
"learning_rate": 1.404977375565611e-06,
|
| 629 |
+
"loss": 0.4402,
|
| 630 |
+
"step": 3800
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 8.710407239819004,
|
| 634 |
+
"grad_norm": 0.049724407494068146,
|
| 635 |
+
"learning_rate": 1.2918552036199098e-06,
|
| 636 |
+
"loss": 0.132,
|
| 637 |
+
"step": 3850
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 8.823529411764707,
|
| 641 |
+
"grad_norm": 0.00581687968224287,
|
| 642 |
+
"learning_rate": 1.1787330316742083e-06,
|
| 643 |
+
"loss": 0.4668,
|
| 644 |
+
"step": 3900
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 8.936651583710407,
|
| 648 |
+
"grad_norm": 0.0993039682507515,
|
| 649 |
+
"learning_rate": 1.065610859728507e-06,
|
| 650 |
+
"loss": 0.9671,
|
| 651 |
+
"step": 3950
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 9.0,
|
| 655 |
+
"eval_f1": 0.6233766233766234,
|
| 656 |
+
"eval_loss": 1.410291075706482,
|
| 657 |
+
"eval_precision": 0.5217391304347826,
|
| 658 |
+
"eval_recall": 0.7741935483870968,
|
| 659 |
+
"eval_runtime": 8.5219,
|
| 660 |
+
"eval_samples_per_second": 207.465,
|
| 661 |
+
"eval_steps_per_second": 13.025,
|
| 662 |
+
"step": 3978
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 9.049773755656108,
|
| 666 |
+
"grad_norm": 0.01741698570549488,
|
| 667 |
+
"learning_rate": 9.524886877828054e-07,
|
| 668 |
+
"loss": 0.0967,
|
| 669 |
+
"step": 4000
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 9.16289592760181,
|
| 673 |
+
"grad_norm": 0.005097161512821913,
|
| 674 |
+
"learning_rate": 8.393665158371041e-07,
|
| 675 |
+
"loss": 0.0198,
|
| 676 |
+
"step": 4050
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 9.276018099547512,
|
| 680 |
+
"grad_norm": 0.005087696015834808,
|
| 681 |
+
"learning_rate": 7.262443438914028e-07,
|
| 682 |
+
"loss": 1.2565,
|
| 683 |
+
"step": 4100
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 9.389140271493213,
|
| 687 |
+
"grad_norm": 0.009162843227386475,
|
| 688 |
+
"learning_rate": 6.131221719457013e-07,
|
| 689 |
+
"loss": 0.0169,
|
| 690 |
+
"step": 4150
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 9.502262443438914,
|
| 694 |
+
"grad_norm": 0.01073523424565792,
|
| 695 |
+
"learning_rate": 5.000000000000001e-07,
|
| 696 |
+
"loss": 0.0172,
|
| 697 |
+
"step": 4200
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 9.615384615384615,
|
| 701 |
+
"grad_norm": 0.004071434028446674,
|
| 702 |
+
"learning_rate": 3.8687782805429867e-07,
|
| 703 |
+
"loss": 0.2068,
|
| 704 |
+
"step": 4250
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 9.728506787330316,
|
| 708 |
+
"grad_norm": 0.0037068105302751064,
|
| 709 |
+
"learning_rate": 2.737556561085973e-07,
|
| 710 |
+
"loss": 0.0234,
|
| 711 |
+
"step": 4300
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 9.841628959276019,
|
| 715 |
+
"grad_norm": 0.0037418717984110117,
|
| 716 |
+
"learning_rate": 1.606334841628959e-07,
|
| 717 |
+
"loss": 0.1985,
|
| 718 |
+
"step": 4350
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 9.95475113122172,
|
| 722 |
+
"grad_norm": 0.0032976313959807158,
|
| 723 |
+
"learning_rate": 4.751131221719457e-08,
|
| 724 |
+
"loss": 0.0612,
|
| 725 |
+
"step": 4400
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 10.0,
|
| 729 |
+
"eval_f1": 0.6233766233766234,
|
| 730 |
+
"eval_loss": 1.7046259641647339,
|
| 731 |
+
"eval_precision": 0.5217391304347826,
|
| 732 |
+
"eval_recall": 0.7741935483870968,
|
| 733 |
+
"eval_runtime": 8.423,
|
| 734 |
+
"eval_samples_per_second": 209.903,
|
| 735 |
+
"eval_steps_per_second": 13.178,
|
| 736 |
+
"step": 4420
|
| 737 |
+
}
|
| 738 |
+
],
|
| 739 |
+
"logging_steps": 50,
|
| 740 |
+
"max_steps": 4420,
|
| 741 |
+
"num_input_tokens_seen": 0,
|
| 742 |
+
"num_train_epochs": 10,
|
| 743 |
+
"save_steps": 500,
|
| 744 |
+
"stateful_callbacks": {
|
| 745 |
+
"TrainerControl": {
|
| 746 |
+
"args": {
|
| 747 |
+
"should_epoch_stop": false,
|
| 748 |
+
"should_evaluate": false,
|
| 749 |
+
"should_log": false,
|
| 750 |
+
"should_save": true,
|
| 751 |
+
"should_training_stop": true
|
| 752 |
+
},
|
| 753 |
+
"attributes": {}
|
| 754 |
+
}
|
| 755 |
+
},
|
| 756 |
+
"total_flos": 2341319516390400.0,
|
| 757 |
+
"train_batch_size": 16,
|
| 758 |
+
"trial_name": null,
|
| 759 |
+
"trial_params": null
|
| 760 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a0c3ebf3358e5bf13ee595da0d31aa1e4508ec78599024e35c7f029521294c6
|
| 3 |
+
size 5713
|