Training in progress, epoch 1
Browse files- model.safetensors +1 -1
- run-2/checkpoint-7200/config.json +125 -0
- run-2/checkpoint-7200/model.safetensors +3 -0
- run-2/checkpoint-7200/optimizer.pt +3 -0
- run-2/checkpoint-7200/rng_state.pth +3 -0
- run-2/checkpoint-7200/scaler.pt +3 -0
- run-2/checkpoint-7200/scheduler.pt +3 -0
- run-2/checkpoint-7200/trainer_state.json +755 -0
- run-2/checkpoint-7200/training_args.bin +3 -0
- run-2/checkpoint-8400/config.json +125 -0
- run-2/checkpoint-8400/model.safetensors +3 -0
- run-2/checkpoint-8400/optimizer.pt +3 -0
- run-2/checkpoint-8400/rng_state.pth +3 -0
- run-2/checkpoint-8400/scaler.pt +3 -0
- run-2/checkpoint-8400/scheduler.pt +3 -0
- run-2/checkpoint-8400/trainer_state.json +873 -0
- run-2/checkpoint-8400/training_args.bin +3 -0
- run-2/checkpoint-9600/config.json +125 -0
- run-2/checkpoint-9600/model.safetensors +3 -0
- run-2/checkpoint-9600/optimizer.pt +3 -0
- run-2/checkpoint-9600/rng_state.pth +3 -0
- run-2/checkpoint-9600/scaler.pt +3 -0
- run-2/checkpoint-9600/scheduler.pt +3 -0
- run-2/checkpoint-9600/trainer_state.json +991 -0
- run-2/checkpoint-9600/training_args.bin +3 -0
- run-3/checkpoint-2399/config.json +125 -0
- run-3/checkpoint-2399/model.safetensors +3 -0
- run-3/checkpoint-2399/optimizer.pt +3 -0
- run-3/checkpoint-2399/rng_state.pth +3 -0
- run-3/checkpoint-2399/scaler.pt +3 -0
- run-3/checkpoint-2399/scheduler.pt +3 -0
- run-3/checkpoint-2399/trainer_state.json +242 -0
- run-3/checkpoint-2399/training_args.bin +3 -0
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 438100144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8110a1a418d11311bbe6626de555c200d3eaaa703654697275995e81a68954e3
|
| 3 |
size 438100144
|
run-2/checkpoint-7200/config.json
ADDED
|
@@ -0,0 +1,125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3",
|
| 16 |
+
"4": "LABEL_4",
|
| 17 |
+
"5": "LABEL_5",
|
| 18 |
+
"6": "LABEL_6",
|
| 19 |
+
"7": "LABEL_7",
|
| 20 |
+
"8": "LABEL_8",
|
| 21 |
+
"9": "LABEL_9",
|
| 22 |
+
"10": "LABEL_10",
|
| 23 |
+
"11": "LABEL_11",
|
| 24 |
+
"12": "LABEL_12",
|
| 25 |
+
"13": "LABEL_13",
|
| 26 |
+
"14": "LABEL_14",
|
| 27 |
+
"15": "LABEL_15",
|
| 28 |
+
"16": "LABEL_16",
|
| 29 |
+
"17": "LABEL_17",
|
| 30 |
+
"18": "LABEL_18",
|
| 31 |
+
"19": "LABEL_19",
|
| 32 |
+
"20": "LABEL_20",
|
| 33 |
+
"21": "LABEL_21",
|
| 34 |
+
"22": "LABEL_22",
|
| 35 |
+
"23": "LABEL_23",
|
| 36 |
+
"24": "LABEL_24",
|
| 37 |
+
"25": "LABEL_25",
|
| 38 |
+
"26": "LABEL_26",
|
| 39 |
+
"27": "LABEL_27",
|
| 40 |
+
"28": "LABEL_28",
|
| 41 |
+
"29": "LABEL_29",
|
| 42 |
+
"30": "LABEL_30",
|
| 43 |
+
"31": "LABEL_31",
|
| 44 |
+
"32": "LABEL_32",
|
| 45 |
+
"33": "LABEL_33",
|
| 46 |
+
"34": "LABEL_34",
|
| 47 |
+
"35": "LABEL_35",
|
| 48 |
+
"36": "LABEL_36",
|
| 49 |
+
"37": "LABEL_37",
|
| 50 |
+
"38": "LABEL_38",
|
| 51 |
+
"39": "LABEL_39",
|
| 52 |
+
"40": "LABEL_40",
|
| 53 |
+
"41": "LABEL_41",
|
| 54 |
+
"42": "LABEL_42",
|
| 55 |
+
"43": "LABEL_43",
|
| 56 |
+
"44": "LABEL_44",
|
| 57 |
+
"45": "LABEL_45",
|
| 58 |
+
"46": "LABEL_46",
|
| 59 |
+
"47": "LABEL_47"
|
| 60 |
+
},
|
| 61 |
+
"initializer_range": 0.02,
|
| 62 |
+
"intermediate_size": 3072,
|
| 63 |
+
"label2id": {
|
| 64 |
+
"LABEL_0": 0,
|
| 65 |
+
"LABEL_1": 1,
|
| 66 |
+
"LABEL_10": 10,
|
| 67 |
+
"LABEL_11": 11,
|
| 68 |
+
"LABEL_12": 12,
|
| 69 |
+
"LABEL_13": 13,
|
| 70 |
+
"LABEL_14": 14,
|
| 71 |
+
"LABEL_15": 15,
|
| 72 |
+
"LABEL_16": 16,
|
| 73 |
+
"LABEL_17": 17,
|
| 74 |
+
"LABEL_18": 18,
|
| 75 |
+
"LABEL_19": 19,
|
| 76 |
+
"LABEL_2": 2,
|
| 77 |
+
"LABEL_20": 20,
|
| 78 |
+
"LABEL_21": 21,
|
| 79 |
+
"LABEL_22": 22,
|
| 80 |
+
"LABEL_23": 23,
|
| 81 |
+
"LABEL_24": 24,
|
| 82 |
+
"LABEL_25": 25,
|
| 83 |
+
"LABEL_26": 26,
|
| 84 |
+
"LABEL_27": 27,
|
| 85 |
+
"LABEL_28": 28,
|
| 86 |
+
"LABEL_29": 29,
|
| 87 |
+
"LABEL_3": 3,
|
| 88 |
+
"LABEL_30": 30,
|
| 89 |
+
"LABEL_31": 31,
|
| 90 |
+
"LABEL_32": 32,
|
| 91 |
+
"LABEL_33": 33,
|
| 92 |
+
"LABEL_34": 34,
|
| 93 |
+
"LABEL_35": 35,
|
| 94 |
+
"LABEL_36": 36,
|
| 95 |
+
"LABEL_37": 37,
|
| 96 |
+
"LABEL_38": 38,
|
| 97 |
+
"LABEL_39": 39,
|
| 98 |
+
"LABEL_4": 4,
|
| 99 |
+
"LABEL_40": 40,
|
| 100 |
+
"LABEL_41": 41,
|
| 101 |
+
"LABEL_42": 42,
|
| 102 |
+
"LABEL_43": 43,
|
| 103 |
+
"LABEL_44": 44,
|
| 104 |
+
"LABEL_45": 45,
|
| 105 |
+
"LABEL_46": 46,
|
| 106 |
+
"LABEL_47": 47,
|
| 107 |
+
"LABEL_5": 5,
|
| 108 |
+
"LABEL_6": 6,
|
| 109 |
+
"LABEL_7": 7,
|
| 110 |
+
"LABEL_8": 8,
|
| 111 |
+
"LABEL_9": 9
|
| 112 |
+
},
|
| 113 |
+
"layer_norm_eps": 1e-12,
|
| 114 |
+
"max_position_embeddings": 512,
|
| 115 |
+
"model_type": "bert",
|
| 116 |
+
"num_attention_heads": 12,
|
| 117 |
+
"num_hidden_layers": 12,
|
| 118 |
+
"pad_token_id": 0,
|
| 119 |
+
"position_embedding_type": "absolute",
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.50.3",
|
| 122 |
+
"type_vocab_size": 2,
|
| 123 |
+
"use_cache": true,
|
| 124 |
+
"vocab_size": 30522
|
| 125 |
+
}
|
run-2/checkpoint-7200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7482134bf65c2abc5bbbb1174696a6279641edcbb0c630dae3fd0fb6f99b979c
|
| 3 |
+
size 438100144
|
run-2/checkpoint-7200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32ab8ccb98d850d06703afbb45b1304fc7cdfa17e68297a4b50187a23aa00c5e
|
| 3 |
+
size 876321402
|
run-2/checkpoint-7200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8950df86f7a6bcf4605ec42c8c63f0262cb16b1bfc5127adb4efa5a0b1b1330d
|
| 3 |
+
size 14244
|
run-2/checkpoint-7200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f5203015529b85590590219a93c1d55fc66b1e8cacd7d5fb766a58ebcd04aaf
|
| 3 |
+
size 988
|
run-2/checkpoint-7200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a1fc8293debaa5b2840c3c3a2a982458a4256cc036e148bf064661ba101cea1
|
| 3 |
+
size 1064
|
run-2/checkpoint-7200/trainer_state.json
ADDED
|
@@ -0,0 +1,755 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7200,
|
| 3 |
+
"best_metric": 0.9362734744251768,
|
| 4 |
+
"best_model_checkpoint": "./results/run-2/checkpoint-7200",
|
| 5 |
+
"epoch": 6.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 7200,
|
| 8 |
+
"is_hyper_param_search": true,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08336807002917883,
|
| 14 |
+
"grad_norm": 1.2878578901290894,
|
| 15 |
+
"learning_rate": 1.4e-05,
|
| 16 |
+
"loss": 1.2644,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16673614005835766,
|
| 21 |
+
"grad_norm": 1.0009047985076904,
|
| 22 |
+
"learning_rate": 2.8e-05,
|
| 23 |
+
"loss": 0.4758,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.25010421008753647,
|
| 28 |
+
"grad_norm": 0.7635347843170166,
|
| 29 |
+
"learning_rate": 4.2e-05,
|
| 30 |
+
"loss": 0.4054,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.3334722801167153,
|
| 35 |
+
"grad_norm": 0.8168604373931885,
|
| 36 |
+
"learning_rate": 5.6e-05,
|
| 37 |
+
"loss": 0.3468,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4168403501458941,
|
| 42 |
+
"grad_norm": 1.1069319248199463,
|
| 43 |
+
"learning_rate": 7e-05,
|
| 44 |
+
"loss": 0.3051,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5002084201750729,
|
| 49 |
+
"grad_norm": 1.015086054801941,
|
| 50 |
+
"learning_rate": 6.962534789124383e-05,
|
| 51 |
+
"loss": 0.3044,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.5835764902042517,
|
| 56 |
+
"grad_norm": 0.8430781364440918,
|
| 57 |
+
"learning_rate": 6.925069578248769e-05,
|
| 58 |
+
"loss": 0.2742,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.6669445602334306,
|
| 63 |
+
"grad_norm": 0.950409471988678,
|
| 64 |
+
"learning_rate": 6.887604367373152e-05,
|
| 65 |
+
"loss": 0.2573,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.7503126302626094,
|
| 70 |
+
"grad_norm": 1.171885371208191,
|
| 71 |
+
"learning_rate": 6.850139156497538e-05,
|
| 72 |
+
"loss": 0.2517,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.8336807002917882,
|
| 77 |
+
"grad_norm": 0.9133914709091187,
|
| 78 |
+
"learning_rate": 6.812673945621922e-05,
|
| 79 |
+
"loss": 0.2435,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.9170487703209671,
|
| 84 |
+
"grad_norm": 0.7939682602882385,
|
| 85 |
+
"learning_rate": 6.775208734746307e-05,
|
| 86 |
+
"loss": 0.2373,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.0,
|
| 91 |
+
"grad_norm": 0.49670639634132385,
|
| 92 |
+
"learning_rate": 6.737743523870691e-05,
|
| 93 |
+
"loss": 0.2363,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.0,
|
| 98 |
+
"eval_administration_accuracy": 0.9302689180737961,
|
| 99 |
+
"eval_administration_f1": 0.9077189587684091,
|
| 100 |
+
"eval_corruption_accuracy": 0.9534083802376485,
|
| 101 |
+
"eval_corruption_f1": 0.9474431988592255,
|
| 102 |
+
"eval_democracy_accuracy": 0.9473629351678132,
|
| 103 |
+
"eval_democracy_f1": 0.9358042521135516,
|
| 104 |
+
"eval_development_accuracy": 0.9094225557640192,
|
| 105 |
+
"eval_development_f1": 0.8937523747915405,
|
| 106 |
+
"eval_economy_accuracy": 0.924119241192412,
|
| 107 |
+
"eval_economy_f1": 0.9156575066932242,
|
| 108 |
+
"eval_education_accuracy": 0.9621638524077548,
|
| 109 |
+
"eval_education_f1": 0.9583809787976314,
|
| 110 |
+
"eval_environment_accuracy": 0.9775901605169898,
|
| 111 |
+
"eval_environment_f1": 0.9754739686963207,
|
| 112 |
+
"eval_instability_accuracy": 0.9422555764019178,
|
| 113 |
+
"eval_instability_f1": 0.9343996801103537,
|
| 114 |
+
"eval_leadership_accuracy": 0.7970606629143214,
|
| 115 |
+
"eval_leadership_f1": 0.7702375495381067,
|
| 116 |
+
"eval_loss": 0.22745274007320404,
|
| 117 |
+
"eval_overall_accuracy": 0.9296261552359112,
|
| 118 |
+
"eval_overall_f1": 0.9199542168750479,
|
| 119 |
+
"eval_race_accuracy": 0.9505941213258287,
|
| 120 |
+
"eval_race_f1": 0.9483610849847235,
|
| 121 |
+
"eval_religion_accuracy": 0.9505941213258287,
|
| 122 |
+
"eval_religion_f1": 0.9489643730683611,
|
| 123 |
+
"eval_runtime": 11.0331,
|
| 124 |
+
"eval_safety_accuracy": 0.9106733375026058,
|
| 125 |
+
"eval_safety_f1": 0.9032566760791263,
|
| 126 |
+
"eval_samples_per_second": 869.566,
|
| 127 |
+
"eval_steps_per_second": 54.382,
|
| 128 |
+
"step": 1200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.0833680700291788,
|
| 132 |
+
"grad_norm": 0.949184238910675,
|
| 133 |
+
"learning_rate": 6.700278312995076e-05,
|
| 134 |
+
"loss": 0.1968,
|
| 135 |
+
"step": 1300
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.1667361400583576,
|
| 139 |
+
"grad_norm": 0.7735128402709961,
|
| 140 |
+
"learning_rate": 6.66281310211946e-05,
|
| 141 |
+
"loss": 0.1924,
|
| 142 |
+
"step": 1400
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.2501042100875366,
|
| 146 |
+
"grad_norm": 1.0187016725540161,
|
| 147 |
+
"learning_rate": 6.625347891243845e-05,
|
| 148 |
+
"loss": 0.1945,
|
| 149 |
+
"step": 1500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3334722801167154,
|
| 153 |
+
"grad_norm": 0.9707878828048706,
|
| 154 |
+
"learning_rate": 6.587882680368229e-05,
|
| 155 |
+
"loss": 0.1931,
|
| 156 |
+
"step": 1600
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.4168403501458942,
|
| 160 |
+
"grad_norm": 1.4602934122085571,
|
| 161 |
+
"learning_rate": 6.550417469492613e-05,
|
| 162 |
+
"loss": 0.1967,
|
| 163 |
+
"step": 1700
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.500208420175073,
|
| 167 |
+
"grad_norm": 1.2449183464050293,
|
| 168 |
+
"learning_rate": 6.512952258616998e-05,
|
| 169 |
+
"loss": 0.1843,
|
| 170 |
+
"step": 1800
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.5835764902042517,
|
| 174 |
+
"grad_norm": 1.4943019151687622,
|
| 175 |
+
"learning_rate": 6.475487047741382e-05,
|
| 176 |
+
"loss": 0.1844,
|
| 177 |
+
"step": 1900
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6669445602334307,
|
| 181 |
+
"grad_norm": 1.8496220111846924,
|
| 182 |
+
"learning_rate": 6.438021836865767e-05,
|
| 183 |
+
"loss": 0.1797,
|
| 184 |
+
"step": 2000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.7503126302626093,
|
| 188 |
+
"grad_norm": 1.0578210353851318,
|
| 189 |
+
"learning_rate": 6.400556625990151e-05,
|
| 190 |
+
"loss": 0.1875,
|
| 191 |
+
"step": 2100
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.8336807002917883,
|
| 195 |
+
"grad_norm": 1.0538480281829834,
|
| 196 |
+
"learning_rate": 6.363091415114536e-05,
|
| 197 |
+
"loss": 0.1685,
|
| 198 |
+
"step": 2200
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.917048770320967,
|
| 202 |
+
"grad_norm": 1.5351332426071167,
|
| 203 |
+
"learning_rate": 6.32562620423892e-05,
|
| 204 |
+
"loss": 0.1744,
|
| 205 |
+
"step": 2300
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.0,
|
| 209 |
+
"grad_norm": 0.9578828811645508,
|
| 210 |
+
"learning_rate": 6.288160993363306e-05,
|
| 211 |
+
"loss": 0.1811,
|
| 212 |
+
"step": 2400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.0,
|
| 216 |
+
"eval_administration_accuracy": 0.9344381905357515,
|
| 217 |
+
"eval_administration_f1": 0.9296143144580357,
|
| 218 |
+
"eval_corruption_accuracy": 0.9604961434229727,
|
| 219 |
+
"eval_corruption_f1": 0.9577803781696625,
|
| 220 |
+
"eval_democracy_accuracy": 0.9534083802376485,
|
| 221 |
+
"eval_democracy_f1": 0.9434755362160818,
|
| 222 |
+
"eval_development_accuracy": 0.9193245778611632,
|
| 223 |
+
"eval_development_f1": 0.9132597892423039,
|
| 224 |
+
"eval_economy_accuracy": 0.931311236189285,
|
| 225 |
+
"eval_economy_f1": 0.92884213698996,
|
| 226 |
+
"eval_education_accuracy": 0.9653950385657702,
|
| 227 |
+
"eval_education_f1": 0.963862144635937,
|
| 228 |
+
"eval_environment_accuracy": 0.979153637690223,
|
| 229 |
+
"eval_environment_f1": 0.9783610154594226,
|
| 230 |
+
"eval_instability_accuracy": 0.9459036898061288,
|
| 231 |
+
"eval_instability_f1": 0.9423381513536221,
|
| 232 |
+
"eval_leadership_accuracy": 0.8387533875338753,
|
| 233 |
+
"eval_leadership_f1": 0.8323519207670947,
|
| 234 |
+
"eval_loss": 0.2031705528497696,
|
| 235 |
+
"eval_overall_accuracy": 0.9372177055103883,
|
| 236 |
+
"eval_overall_f1": 0.9339233416905627,
|
| 237 |
+
"eval_race_accuracy": 0.9516364394413175,
|
| 238 |
+
"eval_race_f1": 0.9498984068182696,
|
| 239 |
+
"eval_religion_accuracy": 0.9558057119032729,
|
| 240 |
+
"eval_religion_f1": 0.9554428854878803,
|
| 241 |
+
"eval_runtime": 11.0378,
|
| 242 |
+
"eval_safety_accuracy": 0.9109860329372524,
|
| 243 |
+
"eval_safety_f1": 0.9118534206884823,
|
| 244 |
+
"eval_samples_per_second": 869.197,
|
| 245 |
+
"eval_steps_per_second": 54.359,
|
| 246 |
+
"step": 2400
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 2.083368070029179,
|
| 250 |
+
"grad_norm": 1.0708891153335571,
|
| 251 |
+
"learning_rate": 6.25069578248769e-05,
|
| 252 |
+
"loss": 0.1338,
|
| 253 |
+
"step": 2500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 2.1667361400583576,
|
| 257 |
+
"grad_norm": 1.0800749063491821,
|
| 258 |
+
"learning_rate": 6.213230571612073e-05,
|
| 259 |
+
"loss": 0.1268,
|
| 260 |
+
"step": 2600
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.2501042100875366,
|
| 264 |
+
"grad_norm": 1.3577998876571655,
|
| 265 |
+
"learning_rate": 6.175765360736459e-05,
|
| 266 |
+
"loss": 0.1226,
|
| 267 |
+
"step": 2700
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.333472280116715,
|
| 271 |
+
"grad_norm": 1.325534701347351,
|
| 272 |
+
"learning_rate": 6.138300149860843e-05,
|
| 273 |
+
"loss": 0.1258,
|
| 274 |
+
"step": 2800
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.416840350145894,
|
| 278 |
+
"grad_norm": 1.7846063375473022,
|
| 279 |
+
"learning_rate": 6.100834938985228e-05,
|
| 280 |
+
"loss": 0.1308,
|
| 281 |
+
"step": 2900
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.500208420175073,
|
| 285 |
+
"grad_norm": 1.61028254032135,
|
| 286 |
+
"learning_rate": 6.0633697281096124e-05,
|
| 287 |
+
"loss": 0.1305,
|
| 288 |
+
"step": 3000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.5835764902042517,
|
| 292 |
+
"grad_norm": 1.2228693962097168,
|
| 293 |
+
"learning_rate": 6.025904517233996e-05,
|
| 294 |
+
"loss": 0.1315,
|
| 295 |
+
"step": 3100
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.6669445602334307,
|
| 299 |
+
"grad_norm": 0.9486532807350159,
|
| 300 |
+
"learning_rate": 5.988439306358381e-05,
|
| 301 |
+
"loss": 0.1337,
|
| 302 |
+
"step": 3200
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.7503126302626093,
|
| 306 |
+
"grad_norm": 1.3199305534362793,
|
| 307 |
+
"learning_rate": 5.9509740954827654e-05,
|
| 308 |
+
"loss": 0.132,
|
| 309 |
+
"step": 3300
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8336807002917883,
|
| 313 |
+
"grad_norm": 1.8821613788604736,
|
| 314 |
+
"learning_rate": 5.91350888460715e-05,
|
| 315 |
+
"loss": 0.1258,
|
| 316 |
+
"step": 3400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 2.9170487703209673,
|
| 320 |
+
"grad_norm": 1.2816129922866821,
|
| 321 |
+
"learning_rate": 5.8760436737315346e-05,
|
| 322 |
+
"loss": 0.1282,
|
| 323 |
+
"step": 3500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.0,
|
| 327 |
+
"grad_norm": 5.737521171569824,
|
| 328 |
+
"learning_rate": 5.838578462855919e-05,
|
| 329 |
+
"loss": 0.1297,
|
| 330 |
+
"step": 3600
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.0,
|
| 334 |
+
"eval_administration_accuracy": 0.9375651448822181,
|
| 335 |
+
"eval_administration_f1": 0.9309689641816119,
|
| 336 |
+
"eval_corruption_accuracy": 0.9451740671252866,
|
| 337 |
+
"eval_corruption_f1": 0.9476001634812502,
|
| 338 |
+
"eval_democracy_accuracy": 0.9508025849489264,
|
| 339 |
+
"eval_democracy_f1": 0.9508783301626133,
|
| 340 |
+
"eval_development_accuracy": 0.921304982280592,
|
| 341 |
+
"eval_development_f1": 0.9145210502797096,
|
| 342 |
+
"eval_economy_accuracy": 0.930373149885345,
|
| 343 |
+
"eval_economy_f1": 0.9257602899924291,
|
| 344 |
+
"eval_education_accuracy": 0.9670627475505524,
|
| 345 |
+
"eval_education_f1": 0.9651765113114141,
|
| 346 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 347 |
+
"eval_environment_f1": 0.9785021531425564,
|
| 348 |
+
"eval_instability_accuracy": 0.9419428809672712,
|
| 349 |
+
"eval_instability_f1": 0.9412425903038227,
|
| 350 |
+
"eval_leadership_accuracy": 0.8284344381905357,
|
| 351 |
+
"eval_leadership_f1": 0.8287745355631846,
|
| 352 |
+
"eval_loss": 0.21378232538700104,
|
| 353 |
+
"eval_overall_accuracy": 0.9360450976304634,
|
| 354 |
+
"eval_overall_f1": 0.9337781089602167,
|
| 355 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 356 |
+
"eval_race_f1": 0.9510004306979282,
|
| 357 |
+
"eval_religion_accuracy": 0.955701480091724,
|
| 358 |
+
"eval_religion_f1": 0.9539633431565642,
|
| 359 |
+
"eval_runtime": 11.1303,
|
| 360 |
+
"eval_safety_accuracy": 0.919533041484261,
|
| 361 |
+
"eval_safety_f1": 0.9169489452495161,
|
| 362 |
+
"eval_samples_per_second": 861.972,
|
| 363 |
+
"eval_steps_per_second": 53.907,
|
| 364 |
+
"step": 3600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 3.083368070029179,
|
| 368 |
+
"grad_norm": 0.7353236079216003,
|
| 369 |
+
"learning_rate": 5.801113251980304e-05,
|
| 370 |
+
"loss": 0.0912,
|
| 371 |
+
"step": 3700
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 3.1667361400583576,
|
| 375 |
+
"grad_norm": 1.2753976583480835,
|
| 376 |
+
"learning_rate": 5.763648041104688e-05,
|
| 377 |
+
"loss": 0.0954,
|
| 378 |
+
"step": 3800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 3.2501042100875366,
|
| 382 |
+
"grad_norm": 1.4709268808364868,
|
| 383 |
+
"learning_rate": 5.726182830229073e-05,
|
| 384 |
+
"loss": 0.0936,
|
| 385 |
+
"step": 3900
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 3.333472280116715,
|
| 389 |
+
"grad_norm": 0.8533086180686951,
|
| 390 |
+
"learning_rate": 5.6887176193534574e-05,
|
| 391 |
+
"loss": 0.0927,
|
| 392 |
+
"step": 4000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.416840350145894,
|
| 396 |
+
"grad_norm": 1.1082295179367065,
|
| 397 |
+
"learning_rate": 5.6512524084778413e-05,
|
| 398 |
+
"loss": 0.091,
|
| 399 |
+
"step": 4100
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 3.500208420175073,
|
| 403 |
+
"grad_norm": 1.3943876028060913,
|
| 404 |
+
"learning_rate": 5.613787197602226e-05,
|
| 405 |
+
"loss": 0.0966,
|
| 406 |
+
"step": 4200
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.5835764902042517,
|
| 410 |
+
"grad_norm": 1.483786940574646,
|
| 411 |
+
"learning_rate": 5.5763219867266105e-05,
|
| 412 |
+
"loss": 0.0965,
|
| 413 |
+
"step": 4300
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 3.6669445602334307,
|
| 417 |
+
"grad_norm": 1.0739402770996094,
|
| 418 |
+
"learning_rate": 5.538856775850995e-05,
|
| 419 |
+
"loss": 0.0988,
|
| 420 |
+
"step": 4400
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 3.7503126302626093,
|
| 424 |
+
"grad_norm": 1.2540950775146484,
|
| 425 |
+
"learning_rate": 5.5013915649753796e-05,
|
| 426 |
+
"loss": 0.0921,
|
| 427 |
+
"step": 4500
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 3.8336807002917883,
|
| 431 |
+
"grad_norm": 1.5991489887237549,
|
| 432 |
+
"learning_rate": 5.463926354099764e-05,
|
| 433 |
+
"loss": 0.1044,
|
| 434 |
+
"step": 4600
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.9170487703209673,
|
| 438 |
+
"grad_norm": 1.71983003616333,
|
| 439 |
+
"learning_rate": 5.426461143224149e-05,
|
| 440 |
+
"loss": 0.0891,
|
| 441 |
+
"step": 4700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 4.0,
|
| 445 |
+
"grad_norm": 0.6225905418395996,
|
| 446 |
+
"learning_rate": 5.3889959323485334e-05,
|
| 447 |
+
"loss": 0.0935,
|
| 448 |
+
"step": 4800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.0,
|
| 452 |
+
"eval_administration_accuracy": 0.9401709401709402,
|
| 453 |
+
"eval_administration_f1": 0.9351367023709064,
|
| 454 |
+
"eval_corruption_accuracy": 0.9568480300187617,
|
| 455 |
+
"eval_corruption_f1": 0.9560031856416787,
|
| 456 |
+
"eval_democracy_accuracy": 0.9563268709610173,
|
| 457 |
+
"eval_democracy_f1": 0.9539147719955986,
|
| 458 |
+
"eval_development_accuracy": 0.9181780279341255,
|
| 459 |
+
"eval_development_f1": 0.9124572322058647,
|
| 460 |
+
"eval_economy_accuracy": 0.929435063581405,
|
| 461 |
+
"eval_economy_f1": 0.9284126365185199,
|
| 462 |
+
"eval_education_accuracy": 0.9672712111736502,
|
| 463 |
+
"eval_education_f1": 0.9657318355589747,
|
| 464 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 465 |
+
"eval_environment_f1": 0.9793120114242748,
|
| 466 |
+
"eval_instability_accuracy": 0.9337085678549093,
|
| 467 |
+
"eval_instability_f1": 0.9360007265571828,
|
| 468 |
+
"eval_leadership_accuracy": 0.8413591828225975,
|
| 469 |
+
"eval_leadership_f1": 0.8351481808317307,
|
| 470 |
+
"eval_loss": 0.22106842696666718,
|
| 471 |
+
"eval_overall_accuracy": 0.9382079077201028,
|
| 472 |
+
"eval_overall_f1": 0.9358746589615409,
|
| 473 |
+
"eval_race_accuracy": 0.9561184073379195,
|
| 474 |
+
"eval_race_f1": 0.9543792719379139,
|
| 475 |
+
"eval_religion_accuracy": 0.9570564936418595,
|
| 476 |
+
"eval_religion_f1": 0.9560165149767178,
|
| 477 |
+
"eval_runtime": 10.9937,
|
| 478 |
+
"eval_safety_accuracy": 0.9215134459036898,
|
| 479 |
+
"eval_safety_f1": 0.9179828375191263,
|
| 480 |
+
"eval_samples_per_second": 872.679,
|
| 481 |
+
"eval_steps_per_second": 54.577,
|
| 482 |
+
"step": 4800
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.083368070029179,
|
| 486 |
+
"grad_norm": 0.6198378801345825,
|
| 487 |
+
"learning_rate": 5.351530721472918e-05,
|
| 488 |
+
"loss": 0.0658,
|
| 489 |
+
"step": 4900
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.166736140058358,
|
| 493 |
+
"grad_norm": 1.2718008756637573,
|
| 494 |
+
"learning_rate": 5.3140655105973025e-05,
|
| 495 |
+
"loss": 0.0636,
|
| 496 |
+
"step": 5000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.250104210087536,
|
| 500 |
+
"grad_norm": 1.476962685585022,
|
| 501 |
+
"learning_rate": 5.2766002997216864e-05,
|
| 502 |
+
"loss": 0.0742,
|
| 503 |
+
"step": 5100
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 4.333472280116715,
|
| 507 |
+
"grad_norm": 1.1485812664031982,
|
| 508 |
+
"learning_rate": 5.239135088846071e-05,
|
| 509 |
+
"loss": 0.0693,
|
| 510 |
+
"step": 5200
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 4.416840350145894,
|
| 514 |
+
"grad_norm": 1.2767895460128784,
|
| 515 |
+
"learning_rate": 5.2016698779704556e-05,
|
| 516 |
+
"loss": 0.0722,
|
| 517 |
+
"step": 5300
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 4.500208420175073,
|
| 521 |
+
"grad_norm": 0.6865003108978271,
|
| 522 |
+
"learning_rate": 5.16420466709484e-05,
|
| 523 |
+
"loss": 0.0701,
|
| 524 |
+
"step": 5400
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.583576490204251,
|
| 528 |
+
"grad_norm": 0.6567270755767822,
|
| 529 |
+
"learning_rate": 5.127114108327981e-05,
|
| 530 |
+
"loss": 0.069,
|
| 531 |
+
"step": 5500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.66694456023343,
|
| 535 |
+
"grad_norm": 1.1187883615493774,
|
| 536 |
+
"learning_rate": 5.0896488974523654e-05,
|
| 537 |
+
"loss": 0.0716,
|
| 538 |
+
"step": 5600
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.750312630262609,
|
| 542 |
+
"grad_norm": 1.3242968320846558,
|
| 543 |
+
"learning_rate": 5.05218368657675e-05,
|
| 544 |
+
"loss": 0.0702,
|
| 545 |
+
"step": 5700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 4.833680700291788,
|
| 549 |
+
"grad_norm": 1.458292841911316,
|
| 550 |
+
"learning_rate": 5.0147184757011346e-05,
|
| 551 |
+
"loss": 0.07,
|
| 552 |
+
"step": 5800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 4.917048770320967,
|
| 556 |
+
"grad_norm": 0.7626857757568359,
|
| 557 |
+
"learning_rate": 4.977253264825519e-05,
|
| 558 |
+
"loss": 0.0733,
|
| 559 |
+
"step": 5900
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 5.0,
|
| 563 |
+
"grad_norm": 0.7239888310432434,
|
| 564 |
+
"learning_rate": 4.939788053949904e-05,
|
| 565 |
+
"loss": 0.0675,
|
| 566 |
+
"step": 6000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.0,
|
| 570 |
+
"eval_administration_accuracy": 0.9306858453199917,
|
| 571 |
+
"eval_administration_f1": 0.9286500450585223,
|
| 572 |
+
"eval_corruption_accuracy": 0.9606003752345216,
|
| 573 |
+
"eval_corruption_f1": 0.9583563972269709,
|
| 574 |
+
"eval_democracy_accuracy": 0.9416301855326246,
|
| 575 |
+
"eval_democracy_f1": 0.9449257041063754,
|
| 576 |
+
"eval_development_accuracy": 0.9198457369189077,
|
| 577 |
+
"eval_development_f1": 0.9154714581737773,
|
| 578 |
+
"eval_economy_accuracy": 0.9344381905357515,
|
| 579 |
+
"eval_economy_f1": 0.9306995601866777,
|
| 580 |
+
"eval_education_accuracy": 0.9673754429851991,
|
| 581 |
+
"eval_education_f1": 0.9648349073741065,
|
| 582 |
+
"eval_environment_accuracy": 0.9803001876172608,
|
| 583 |
+
"eval_environment_f1": 0.9792105415103222,
|
| 584 |
+
"eval_instability_accuracy": 0.944757139879091,
|
| 585 |
+
"eval_instability_f1": 0.9391731690067155,
|
| 586 |
+
"eval_leadership_accuracy": 0.8202001250781739,
|
| 587 |
+
"eval_leadership_f1": 0.8225964840638132,
|
| 588 |
+
"eval_loss": 0.2460281401872635,
|
| 589 |
+
"eval_overall_accuracy": 0.9360364116461678,
|
| 590 |
+
"eval_overall_f1": 0.9339890155373726,
|
| 591 |
+
"eval_race_accuracy": 0.9508025849489264,
|
| 592 |
+
"eval_race_f1": 0.9501627076296991,
|
| 593 |
+
"eval_religion_accuracy": 0.9597665207421305,
|
| 594 |
+
"eval_religion_f1": 0.9575416931106416,
|
| 595 |
+
"eval_runtime": 10.9751,
|
| 596 |
+
"eval_safety_accuracy": 0.9220346049614342,
|
| 597 |
+
"eval_safety_f1": 0.916245519000848,
|
| 598 |
+
"eval_samples_per_second": 874.157,
|
| 599 |
+
"eval_steps_per_second": 54.669,
|
| 600 |
+
"step": 6000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 5.083368070029179,
|
| 604 |
+
"grad_norm": 1.2741373777389526,
|
| 605 |
+
"learning_rate": 4.902322843074288e-05,
|
| 606 |
+
"loss": 0.0551,
|
| 607 |
+
"step": 6100
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 5.166736140058358,
|
| 611 |
+
"grad_norm": 0.5817021727561951,
|
| 612 |
+
"learning_rate": 4.864857632198673e-05,
|
| 613 |
+
"loss": 0.0505,
|
| 614 |
+
"step": 6200
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 5.250104210087536,
|
| 618 |
+
"grad_norm": 1.0689315795898438,
|
| 619 |
+
"learning_rate": 4.827392421323057e-05,
|
| 620 |
+
"loss": 0.049,
|
| 621 |
+
"step": 6300
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 5.333472280116715,
|
| 625 |
+
"grad_norm": 1.5673719644546509,
|
| 626 |
+
"learning_rate": 4.789927210447441e-05,
|
| 627 |
+
"loss": 0.0554,
|
| 628 |
+
"step": 6400
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 5.416840350145894,
|
| 632 |
+
"grad_norm": 1.1274124383926392,
|
| 633 |
+
"learning_rate": 4.752461999571826e-05,
|
| 634 |
+
"loss": 0.0523,
|
| 635 |
+
"step": 6500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 5.500208420175073,
|
| 639 |
+
"grad_norm": 0.836388349533081,
|
| 640 |
+
"learning_rate": 4.7149967886962105e-05,
|
| 641 |
+
"loss": 0.0516,
|
| 642 |
+
"step": 6600
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 5.583576490204251,
|
| 646 |
+
"grad_norm": 1.7012029886245728,
|
| 647 |
+
"learning_rate": 4.677531577820595e-05,
|
| 648 |
+
"loss": 0.0559,
|
| 649 |
+
"step": 6700
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 5.66694456023343,
|
| 653 |
+
"grad_norm": 1.0244345664978027,
|
| 654 |
+
"learning_rate": 4.6400663669449796e-05,
|
| 655 |
+
"loss": 0.0552,
|
| 656 |
+
"step": 6800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.750312630262609,
|
| 660 |
+
"grad_norm": 1.2528122663497925,
|
| 661 |
+
"learning_rate": 4.602601156069364e-05,
|
| 662 |
+
"loss": 0.0508,
|
| 663 |
+
"step": 6900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.833680700291788,
|
| 667 |
+
"grad_norm": 1.044662594795227,
|
| 668 |
+
"learning_rate": 4.565135945193749e-05,
|
| 669 |
+
"loss": 0.0531,
|
| 670 |
+
"step": 7000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 5.917048770320967,
|
| 674 |
+
"grad_norm": 0.8813854455947876,
|
| 675 |
+
"learning_rate": 4.5276707343181333e-05,
|
| 676 |
+
"loss": 0.0482,
|
| 677 |
+
"step": 7100
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 6.0,
|
| 681 |
+
"grad_norm": 1.4556822776794434,
|
| 682 |
+
"learning_rate": 4.490205523442518e-05,
|
| 683 |
+
"loss": 0.0516,
|
| 684 |
+
"step": 7200
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 6.0,
|
| 688 |
+
"eval_administration_accuracy": 0.9356889722743381,
|
| 689 |
+
"eval_administration_f1": 0.9317820245044639,
|
| 690 |
+
"eval_corruption_accuracy": 0.9579945799457995,
|
| 691 |
+
"eval_corruption_f1": 0.9568343721711066,
|
| 692 |
+
"eval_democracy_accuracy": 0.9589326662497394,
|
| 693 |
+
"eval_democracy_f1": 0.95607216308818,
|
| 694 |
+
"eval_development_accuracy": 0.9168230143839899,
|
| 695 |
+
"eval_development_f1": 0.9143401221807914,
|
| 696 |
+
"eval_economy_accuracy": 0.9308943089430894,
|
| 697 |
+
"eval_economy_f1": 0.9294738147048666,
|
| 698 |
+
"eval_education_accuracy": 0.9669585157390036,
|
| 699 |
+
"eval_education_f1": 0.965464450546169,
|
| 700 |
+
"eval_environment_accuracy": 0.9786324786324786,
|
| 701 |
+
"eval_environment_f1": 0.977778980162068,
|
| 702 |
+
"eval_instability_accuracy": 0.9441317490097978,
|
| 703 |
+
"eval_instability_f1": 0.9395705042517813,
|
| 704 |
+
"eval_leadership_accuracy": 0.8412549510110485,
|
| 705 |
+
"eval_leadership_f1": 0.8395230039371011,
|
| 706 |
+
"eval_loss": 0.26496145129203796,
|
| 707 |
+
"eval_overall_accuracy": 0.938372941421722,
|
| 708 |
+
"eval_overall_f1": 0.9362734744251768,
|
| 709 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 710 |
+
"eval_race_f1": 0.9519598766506976,
|
| 711 |
+
"eval_religion_accuracy": 0.9579945799457995,
|
| 712 |
+
"eval_religion_f1": 0.9568944771373694,
|
| 713 |
+
"eval_runtime": 10.8844,
|
| 714 |
+
"eval_safety_accuracy": 0.9170314780070877,
|
| 715 |
+
"eval_safety_f1": 0.915587903767528,
|
| 716 |
+
"eval_samples_per_second": 881.445,
|
| 717 |
+
"eval_steps_per_second": 55.125,
|
| 718 |
+
"step": 7200
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 100,
|
| 722 |
+
"max_steps": 19184,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 16,
|
| 725 |
+
"save_steps": 500,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"EarlyStoppingCallback": {
|
| 728 |
+
"args": {
|
| 729 |
+
"early_stopping_patience": 2,
|
| 730 |
+
"early_stopping_threshold": 0.0
|
| 731 |
+
},
|
| 732 |
+
"attributes": {
|
| 733 |
+
"early_stopping_patience_counter": 0
|
| 734 |
+
}
|
| 735 |
+
},
|
| 736 |
+
"TrainerControl": {
|
| 737 |
+
"args": {
|
| 738 |
+
"should_epoch_stop": false,
|
| 739 |
+
"should_evaluate": false,
|
| 740 |
+
"should_log": false,
|
| 741 |
+
"should_save": true,
|
| 742 |
+
"should_training_stop": false
|
| 743 |
+
},
|
| 744 |
+
"attributes": {}
|
| 745 |
+
}
|
| 746 |
+
},
|
| 747 |
+
"total_flos": 6.060160356227482e+16,
|
| 748 |
+
"train_batch_size": 16,
|
| 749 |
+
"trial_name": null,
|
| 750 |
+
"trial_params": {
|
| 751 |
+
"gradient_accumulation_steps": 2,
|
| 752 |
+
"learning_rate": 7e-05,
|
| 753 |
+
"num_train_epochs": 16
|
| 754 |
+
}
|
| 755 |
+
}
|
run-2/checkpoint-7200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a417688ba3a38960888d219af94bafd1d9551df9005996513f2ac72dfc76c145
|
| 3 |
+
size 5368
|
run-2/checkpoint-8400/config.json
ADDED
|
@@ -0,0 +1,125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3",
|
| 16 |
+
"4": "LABEL_4",
|
| 17 |
+
"5": "LABEL_5",
|
| 18 |
+
"6": "LABEL_6",
|
| 19 |
+
"7": "LABEL_7",
|
| 20 |
+
"8": "LABEL_8",
|
| 21 |
+
"9": "LABEL_9",
|
| 22 |
+
"10": "LABEL_10",
|
| 23 |
+
"11": "LABEL_11",
|
| 24 |
+
"12": "LABEL_12",
|
| 25 |
+
"13": "LABEL_13",
|
| 26 |
+
"14": "LABEL_14",
|
| 27 |
+
"15": "LABEL_15",
|
| 28 |
+
"16": "LABEL_16",
|
| 29 |
+
"17": "LABEL_17",
|
| 30 |
+
"18": "LABEL_18",
|
| 31 |
+
"19": "LABEL_19",
|
| 32 |
+
"20": "LABEL_20",
|
| 33 |
+
"21": "LABEL_21",
|
| 34 |
+
"22": "LABEL_22",
|
| 35 |
+
"23": "LABEL_23",
|
| 36 |
+
"24": "LABEL_24",
|
| 37 |
+
"25": "LABEL_25",
|
| 38 |
+
"26": "LABEL_26",
|
| 39 |
+
"27": "LABEL_27",
|
| 40 |
+
"28": "LABEL_28",
|
| 41 |
+
"29": "LABEL_29",
|
| 42 |
+
"30": "LABEL_30",
|
| 43 |
+
"31": "LABEL_31",
|
| 44 |
+
"32": "LABEL_32",
|
| 45 |
+
"33": "LABEL_33",
|
| 46 |
+
"34": "LABEL_34",
|
| 47 |
+
"35": "LABEL_35",
|
| 48 |
+
"36": "LABEL_36",
|
| 49 |
+
"37": "LABEL_37",
|
| 50 |
+
"38": "LABEL_38",
|
| 51 |
+
"39": "LABEL_39",
|
| 52 |
+
"40": "LABEL_40",
|
| 53 |
+
"41": "LABEL_41",
|
| 54 |
+
"42": "LABEL_42",
|
| 55 |
+
"43": "LABEL_43",
|
| 56 |
+
"44": "LABEL_44",
|
| 57 |
+
"45": "LABEL_45",
|
| 58 |
+
"46": "LABEL_46",
|
| 59 |
+
"47": "LABEL_47"
|
| 60 |
+
},
|
| 61 |
+
"initializer_range": 0.02,
|
| 62 |
+
"intermediate_size": 3072,
|
| 63 |
+
"label2id": {
|
| 64 |
+
"LABEL_0": 0,
|
| 65 |
+
"LABEL_1": 1,
|
| 66 |
+
"LABEL_10": 10,
|
| 67 |
+
"LABEL_11": 11,
|
| 68 |
+
"LABEL_12": 12,
|
| 69 |
+
"LABEL_13": 13,
|
| 70 |
+
"LABEL_14": 14,
|
| 71 |
+
"LABEL_15": 15,
|
| 72 |
+
"LABEL_16": 16,
|
| 73 |
+
"LABEL_17": 17,
|
| 74 |
+
"LABEL_18": 18,
|
| 75 |
+
"LABEL_19": 19,
|
| 76 |
+
"LABEL_2": 2,
|
| 77 |
+
"LABEL_20": 20,
|
| 78 |
+
"LABEL_21": 21,
|
| 79 |
+
"LABEL_22": 22,
|
| 80 |
+
"LABEL_23": 23,
|
| 81 |
+
"LABEL_24": 24,
|
| 82 |
+
"LABEL_25": 25,
|
| 83 |
+
"LABEL_26": 26,
|
| 84 |
+
"LABEL_27": 27,
|
| 85 |
+
"LABEL_28": 28,
|
| 86 |
+
"LABEL_29": 29,
|
| 87 |
+
"LABEL_3": 3,
|
| 88 |
+
"LABEL_30": 30,
|
| 89 |
+
"LABEL_31": 31,
|
| 90 |
+
"LABEL_32": 32,
|
| 91 |
+
"LABEL_33": 33,
|
| 92 |
+
"LABEL_34": 34,
|
| 93 |
+
"LABEL_35": 35,
|
| 94 |
+
"LABEL_36": 36,
|
| 95 |
+
"LABEL_37": 37,
|
| 96 |
+
"LABEL_38": 38,
|
| 97 |
+
"LABEL_39": 39,
|
| 98 |
+
"LABEL_4": 4,
|
| 99 |
+
"LABEL_40": 40,
|
| 100 |
+
"LABEL_41": 41,
|
| 101 |
+
"LABEL_42": 42,
|
| 102 |
+
"LABEL_43": 43,
|
| 103 |
+
"LABEL_44": 44,
|
| 104 |
+
"LABEL_45": 45,
|
| 105 |
+
"LABEL_46": 46,
|
| 106 |
+
"LABEL_47": 47,
|
| 107 |
+
"LABEL_5": 5,
|
| 108 |
+
"LABEL_6": 6,
|
| 109 |
+
"LABEL_7": 7,
|
| 110 |
+
"LABEL_8": 8,
|
| 111 |
+
"LABEL_9": 9
|
| 112 |
+
},
|
| 113 |
+
"layer_norm_eps": 1e-12,
|
| 114 |
+
"max_position_embeddings": 512,
|
| 115 |
+
"model_type": "bert",
|
| 116 |
+
"num_attention_heads": 12,
|
| 117 |
+
"num_hidden_layers": 12,
|
| 118 |
+
"pad_token_id": 0,
|
| 119 |
+
"position_embedding_type": "absolute",
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.50.3",
|
| 122 |
+
"type_vocab_size": 2,
|
| 123 |
+
"use_cache": true,
|
| 124 |
+
"vocab_size": 30522
|
| 125 |
+
}
|
run-2/checkpoint-8400/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e236524a027f39185d6dbad0b11a890f9d7b892dd571c6bf435c5689dc659bbc
|
| 3 |
+
size 438100144
|
run-2/checkpoint-8400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14342f98f6f49854e4518a4116c7b665a07ecdf27cf30a4ce7c7116cd745cc0a
|
| 3 |
+
size 876321402
|
run-2/checkpoint-8400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:428bd6bed852b4f8932455db9c461eeba79d3efe1fc880d20395baef552a0694
|
| 3 |
+
size 14244
|
run-2/checkpoint-8400/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:711e5f6a2fb64d3c723f5401defc7a49135eaa677c76fcf84d1f5e375e67c5c3
|
| 3 |
+
size 988
|
run-2/checkpoint-8400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2e44344a0e4d0611737d4f73f962224bf500c51bbb326a8091f17f60b1bcfda
|
| 3 |
+
size 1064
|
run-2/checkpoint-8400/trainer_state.json
ADDED
|
@@ -0,0 +1,873 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7200,
|
| 3 |
+
"best_metric": 0.9362734744251768,
|
| 4 |
+
"best_model_checkpoint": "./results/run-2/checkpoint-7200",
|
| 5 |
+
"epoch": 7.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 8400,
|
| 8 |
+
"is_hyper_param_search": true,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08336807002917883,
|
| 14 |
+
"grad_norm": 1.2878578901290894,
|
| 15 |
+
"learning_rate": 1.4e-05,
|
| 16 |
+
"loss": 1.2644,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16673614005835766,
|
| 21 |
+
"grad_norm": 1.0009047985076904,
|
| 22 |
+
"learning_rate": 2.8e-05,
|
| 23 |
+
"loss": 0.4758,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.25010421008753647,
|
| 28 |
+
"grad_norm": 0.7635347843170166,
|
| 29 |
+
"learning_rate": 4.2e-05,
|
| 30 |
+
"loss": 0.4054,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.3334722801167153,
|
| 35 |
+
"grad_norm": 0.8168604373931885,
|
| 36 |
+
"learning_rate": 5.6e-05,
|
| 37 |
+
"loss": 0.3468,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4168403501458941,
|
| 42 |
+
"grad_norm": 1.1069319248199463,
|
| 43 |
+
"learning_rate": 7e-05,
|
| 44 |
+
"loss": 0.3051,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5002084201750729,
|
| 49 |
+
"grad_norm": 1.015086054801941,
|
| 50 |
+
"learning_rate": 6.962534789124383e-05,
|
| 51 |
+
"loss": 0.3044,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.5835764902042517,
|
| 56 |
+
"grad_norm": 0.8430781364440918,
|
| 57 |
+
"learning_rate": 6.925069578248769e-05,
|
| 58 |
+
"loss": 0.2742,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.6669445602334306,
|
| 63 |
+
"grad_norm": 0.950409471988678,
|
| 64 |
+
"learning_rate": 6.887604367373152e-05,
|
| 65 |
+
"loss": 0.2573,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.7503126302626094,
|
| 70 |
+
"grad_norm": 1.171885371208191,
|
| 71 |
+
"learning_rate": 6.850139156497538e-05,
|
| 72 |
+
"loss": 0.2517,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.8336807002917882,
|
| 77 |
+
"grad_norm": 0.9133914709091187,
|
| 78 |
+
"learning_rate": 6.812673945621922e-05,
|
| 79 |
+
"loss": 0.2435,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.9170487703209671,
|
| 84 |
+
"grad_norm": 0.7939682602882385,
|
| 85 |
+
"learning_rate": 6.775208734746307e-05,
|
| 86 |
+
"loss": 0.2373,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.0,
|
| 91 |
+
"grad_norm": 0.49670639634132385,
|
| 92 |
+
"learning_rate": 6.737743523870691e-05,
|
| 93 |
+
"loss": 0.2363,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.0,
|
| 98 |
+
"eval_administration_accuracy": 0.9302689180737961,
|
| 99 |
+
"eval_administration_f1": 0.9077189587684091,
|
| 100 |
+
"eval_corruption_accuracy": 0.9534083802376485,
|
| 101 |
+
"eval_corruption_f1": 0.9474431988592255,
|
| 102 |
+
"eval_democracy_accuracy": 0.9473629351678132,
|
| 103 |
+
"eval_democracy_f1": 0.9358042521135516,
|
| 104 |
+
"eval_development_accuracy": 0.9094225557640192,
|
| 105 |
+
"eval_development_f1": 0.8937523747915405,
|
| 106 |
+
"eval_economy_accuracy": 0.924119241192412,
|
| 107 |
+
"eval_economy_f1": 0.9156575066932242,
|
| 108 |
+
"eval_education_accuracy": 0.9621638524077548,
|
| 109 |
+
"eval_education_f1": 0.9583809787976314,
|
| 110 |
+
"eval_environment_accuracy": 0.9775901605169898,
|
| 111 |
+
"eval_environment_f1": 0.9754739686963207,
|
| 112 |
+
"eval_instability_accuracy": 0.9422555764019178,
|
| 113 |
+
"eval_instability_f1": 0.9343996801103537,
|
| 114 |
+
"eval_leadership_accuracy": 0.7970606629143214,
|
| 115 |
+
"eval_leadership_f1": 0.7702375495381067,
|
| 116 |
+
"eval_loss": 0.22745274007320404,
|
| 117 |
+
"eval_overall_accuracy": 0.9296261552359112,
|
| 118 |
+
"eval_overall_f1": 0.9199542168750479,
|
| 119 |
+
"eval_race_accuracy": 0.9505941213258287,
|
| 120 |
+
"eval_race_f1": 0.9483610849847235,
|
| 121 |
+
"eval_religion_accuracy": 0.9505941213258287,
|
| 122 |
+
"eval_religion_f1": 0.9489643730683611,
|
| 123 |
+
"eval_runtime": 11.0331,
|
| 124 |
+
"eval_safety_accuracy": 0.9106733375026058,
|
| 125 |
+
"eval_safety_f1": 0.9032566760791263,
|
| 126 |
+
"eval_samples_per_second": 869.566,
|
| 127 |
+
"eval_steps_per_second": 54.382,
|
| 128 |
+
"step": 1200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.0833680700291788,
|
| 132 |
+
"grad_norm": 0.949184238910675,
|
| 133 |
+
"learning_rate": 6.700278312995076e-05,
|
| 134 |
+
"loss": 0.1968,
|
| 135 |
+
"step": 1300
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.1667361400583576,
|
| 139 |
+
"grad_norm": 0.7735128402709961,
|
| 140 |
+
"learning_rate": 6.66281310211946e-05,
|
| 141 |
+
"loss": 0.1924,
|
| 142 |
+
"step": 1400
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.2501042100875366,
|
| 146 |
+
"grad_norm": 1.0187016725540161,
|
| 147 |
+
"learning_rate": 6.625347891243845e-05,
|
| 148 |
+
"loss": 0.1945,
|
| 149 |
+
"step": 1500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3334722801167154,
|
| 153 |
+
"grad_norm": 0.9707878828048706,
|
| 154 |
+
"learning_rate": 6.587882680368229e-05,
|
| 155 |
+
"loss": 0.1931,
|
| 156 |
+
"step": 1600
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.4168403501458942,
|
| 160 |
+
"grad_norm": 1.4602934122085571,
|
| 161 |
+
"learning_rate": 6.550417469492613e-05,
|
| 162 |
+
"loss": 0.1967,
|
| 163 |
+
"step": 1700
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.500208420175073,
|
| 167 |
+
"grad_norm": 1.2449183464050293,
|
| 168 |
+
"learning_rate": 6.512952258616998e-05,
|
| 169 |
+
"loss": 0.1843,
|
| 170 |
+
"step": 1800
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.5835764902042517,
|
| 174 |
+
"grad_norm": 1.4943019151687622,
|
| 175 |
+
"learning_rate": 6.475487047741382e-05,
|
| 176 |
+
"loss": 0.1844,
|
| 177 |
+
"step": 1900
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6669445602334307,
|
| 181 |
+
"grad_norm": 1.8496220111846924,
|
| 182 |
+
"learning_rate": 6.438021836865767e-05,
|
| 183 |
+
"loss": 0.1797,
|
| 184 |
+
"step": 2000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.7503126302626093,
|
| 188 |
+
"grad_norm": 1.0578210353851318,
|
| 189 |
+
"learning_rate": 6.400556625990151e-05,
|
| 190 |
+
"loss": 0.1875,
|
| 191 |
+
"step": 2100
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.8336807002917883,
|
| 195 |
+
"grad_norm": 1.0538480281829834,
|
| 196 |
+
"learning_rate": 6.363091415114536e-05,
|
| 197 |
+
"loss": 0.1685,
|
| 198 |
+
"step": 2200
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.917048770320967,
|
| 202 |
+
"grad_norm": 1.5351332426071167,
|
| 203 |
+
"learning_rate": 6.32562620423892e-05,
|
| 204 |
+
"loss": 0.1744,
|
| 205 |
+
"step": 2300
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.0,
|
| 209 |
+
"grad_norm": 0.9578828811645508,
|
| 210 |
+
"learning_rate": 6.288160993363306e-05,
|
| 211 |
+
"loss": 0.1811,
|
| 212 |
+
"step": 2400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.0,
|
| 216 |
+
"eval_administration_accuracy": 0.9344381905357515,
|
| 217 |
+
"eval_administration_f1": 0.9296143144580357,
|
| 218 |
+
"eval_corruption_accuracy": 0.9604961434229727,
|
| 219 |
+
"eval_corruption_f1": 0.9577803781696625,
|
| 220 |
+
"eval_democracy_accuracy": 0.9534083802376485,
|
| 221 |
+
"eval_democracy_f1": 0.9434755362160818,
|
| 222 |
+
"eval_development_accuracy": 0.9193245778611632,
|
| 223 |
+
"eval_development_f1": 0.9132597892423039,
|
| 224 |
+
"eval_economy_accuracy": 0.931311236189285,
|
| 225 |
+
"eval_economy_f1": 0.92884213698996,
|
| 226 |
+
"eval_education_accuracy": 0.9653950385657702,
|
| 227 |
+
"eval_education_f1": 0.963862144635937,
|
| 228 |
+
"eval_environment_accuracy": 0.979153637690223,
|
| 229 |
+
"eval_environment_f1": 0.9783610154594226,
|
| 230 |
+
"eval_instability_accuracy": 0.9459036898061288,
|
| 231 |
+
"eval_instability_f1": 0.9423381513536221,
|
| 232 |
+
"eval_leadership_accuracy": 0.8387533875338753,
|
| 233 |
+
"eval_leadership_f1": 0.8323519207670947,
|
| 234 |
+
"eval_loss": 0.2031705528497696,
|
| 235 |
+
"eval_overall_accuracy": 0.9372177055103883,
|
| 236 |
+
"eval_overall_f1": 0.9339233416905627,
|
| 237 |
+
"eval_race_accuracy": 0.9516364394413175,
|
| 238 |
+
"eval_race_f1": 0.9498984068182696,
|
| 239 |
+
"eval_religion_accuracy": 0.9558057119032729,
|
| 240 |
+
"eval_religion_f1": 0.9554428854878803,
|
| 241 |
+
"eval_runtime": 11.0378,
|
| 242 |
+
"eval_safety_accuracy": 0.9109860329372524,
|
| 243 |
+
"eval_safety_f1": 0.9118534206884823,
|
| 244 |
+
"eval_samples_per_second": 869.197,
|
| 245 |
+
"eval_steps_per_second": 54.359,
|
| 246 |
+
"step": 2400
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 2.083368070029179,
|
| 250 |
+
"grad_norm": 1.0708891153335571,
|
| 251 |
+
"learning_rate": 6.25069578248769e-05,
|
| 252 |
+
"loss": 0.1338,
|
| 253 |
+
"step": 2500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 2.1667361400583576,
|
| 257 |
+
"grad_norm": 1.0800749063491821,
|
| 258 |
+
"learning_rate": 6.213230571612073e-05,
|
| 259 |
+
"loss": 0.1268,
|
| 260 |
+
"step": 2600
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.2501042100875366,
|
| 264 |
+
"grad_norm": 1.3577998876571655,
|
| 265 |
+
"learning_rate": 6.175765360736459e-05,
|
| 266 |
+
"loss": 0.1226,
|
| 267 |
+
"step": 2700
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.333472280116715,
|
| 271 |
+
"grad_norm": 1.325534701347351,
|
| 272 |
+
"learning_rate": 6.138300149860843e-05,
|
| 273 |
+
"loss": 0.1258,
|
| 274 |
+
"step": 2800
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.416840350145894,
|
| 278 |
+
"grad_norm": 1.7846063375473022,
|
| 279 |
+
"learning_rate": 6.100834938985228e-05,
|
| 280 |
+
"loss": 0.1308,
|
| 281 |
+
"step": 2900
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.500208420175073,
|
| 285 |
+
"grad_norm": 1.61028254032135,
|
| 286 |
+
"learning_rate": 6.0633697281096124e-05,
|
| 287 |
+
"loss": 0.1305,
|
| 288 |
+
"step": 3000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.5835764902042517,
|
| 292 |
+
"grad_norm": 1.2228693962097168,
|
| 293 |
+
"learning_rate": 6.025904517233996e-05,
|
| 294 |
+
"loss": 0.1315,
|
| 295 |
+
"step": 3100
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.6669445602334307,
|
| 299 |
+
"grad_norm": 0.9486532807350159,
|
| 300 |
+
"learning_rate": 5.988439306358381e-05,
|
| 301 |
+
"loss": 0.1337,
|
| 302 |
+
"step": 3200
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.7503126302626093,
|
| 306 |
+
"grad_norm": 1.3199305534362793,
|
| 307 |
+
"learning_rate": 5.9509740954827654e-05,
|
| 308 |
+
"loss": 0.132,
|
| 309 |
+
"step": 3300
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8336807002917883,
|
| 313 |
+
"grad_norm": 1.8821613788604736,
|
| 314 |
+
"learning_rate": 5.91350888460715e-05,
|
| 315 |
+
"loss": 0.1258,
|
| 316 |
+
"step": 3400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 2.9170487703209673,
|
| 320 |
+
"grad_norm": 1.2816129922866821,
|
| 321 |
+
"learning_rate": 5.8760436737315346e-05,
|
| 322 |
+
"loss": 0.1282,
|
| 323 |
+
"step": 3500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.0,
|
| 327 |
+
"grad_norm": 5.737521171569824,
|
| 328 |
+
"learning_rate": 5.838578462855919e-05,
|
| 329 |
+
"loss": 0.1297,
|
| 330 |
+
"step": 3600
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.0,
|
| 334 |
+
"eval_administration_accuracy": 0.9375651448822181,
|
| 335 |
+
"eval_administration_f1": 0.9309689641816119,
|
| 336 |
+
"eval_corruption_accuracy": 0.9451740671252866,
|
| 337 |
+
"eval_corruption_f1": 0.9476001634812502,
|
| 338 |
+
"eval_democracy_accuracy": 0.9508025849489264,
|
| 339 |
+
"eval_democracy_f1": 0.9508783301626133,
|
| 340 |
+
"eval_development_accuracy": 0.921304982280592,
|
| 341 |
+
"eval_development_f1": 0.9145210502797096,
|
| 342 |
+
"eval_economy_accuracy": 0.930373149885345,
|
| 343 |
+
"eval_economy_f1": 0.9257602899924291,
|
| 344 |
+
"eval_education_accuracy": 0.9670627475505524,
|
| 345 |
+
"eval_education_f1": 0.9651765113114141,
|
| 346 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 347 |
+
"eval_environment_f1": 0.9785021531425564,
|
| 348 |
+
"eval_instability_accuracy": 0.9419428809672712,
|
| 349 |
+
"eval_instability_f1": 0.9412425903038227,
|
| 350 |
+
"eval_leadership_accuracy": 0.8284344381905357,
|
| 351 |
+
"eval_leadership_f1": 0.8287745355631846,
|
| 352 |
+
"eval_loss": 0.21378232538700104,
|
| 353 |
+
"eval_overall_accuracy": 0.9360450976304634,
|
| 354 |
+
"eval_overall_f1": 0.9337781089602167,
|
| 355 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 356 |
+
"eval_race_f1": 0.9510004306979282,
|
| 357 |
+
"eval_religion_accuracy": 0.955701480091724,
|
| 358 |
+
"eval_religion_f1": 0.9539633431565642,
|
| 359 |
+
"eval_runtime": 11.1303,
|
| 360 |
+
"eval_safety_accuracy": 0.919533041484261,
|
| 361 |
+
"eval_safety_f1": 0.9169489452495161,
|
| 362 |
+
"eval_samples_per_second": 861.972,
|
| 363 |
+
"eval_steps_per_second": 53.907,
|
| 364 |
+
"step": 3600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 3.083368070029179,
|
| 368 |
+
"grad_norm": 0.7353236079216003,
|
| 369 |
+
"learning_rate": 5.801113251980304e-05,
|
| 370 |
+
"loss": 0.0912,
|
| 371 |
+
"step": 3700
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 3.1667361400583576,
|
| 375 |
+
"grad_norm": 1.2753976583480835,
|
| 376 |
+
"learning_rate": 5.763648041104688e-05,
|
| 377 |
+
"loss": 0.0954,
|
| 378 |
+
"step": 3800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 3.2501042100875366,
|
| 382 |
+
"grad_norm": 1.4709268808364868,
|
| 383 |
+
"learning_rate": 5.726182830229073e-05,
|
| 384 |
+
"loss": 0.0936,
|
| 385 |
+
"step": 3900
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 3.333472280116715,
|
| 389 |
+
"grad_norm": 0.8533086180686951,
|
| 390 |
+
"learning_rate": 5.6887176193534574e-05,
|
| 391 |
+
"loss": 0.0927,
|
| 392 |
+
"step": 4000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.416840350145894,
|
| 396 |
+
"grad_norm": 1.1082295179367065,
|
| 397 |
+
"learning_rate": 5.6512524084778413e-05,
|
| 398 |
+
"loss": 0.091,
|
| 399 |
+
"step": 4100
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 3.500208420175073,
|
| 403 |
+
"grad_norm": 1.3943876028060913,
|
| 404 |
+
"learning_rate": 5.613787197602226e-05,
|
| 405 |
+
"loss": 0.0966,
|
| 406 |
+
"step": 4200
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.5835764902042517,
|
| 410 |
+
"grad_norm": 1.483786940574646,
|
| 411 |
+
"learning_rate": 5.5763219867266105e-05,
|
| 412 |
+
"loss": 0.0965,
|
| 413 |
+
"step": 4300
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 3.6669445602334307,
|
| 417 |
+
"grad_norm": 1.0739402770996094,
|
| 418 |
+
"learning_rate": 5.538856775850995e-05,
|
| 419 |
+
"loss": 0.0988,
|
| 420 |
+
"step": 4400
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 3.7503126302626093,
|
| 424 |
+
"grad_norm": 1.2540950775146484,
|
| 425 |
+
"learning_rate": 5.5013915649753796e-05,
|
| 426 |
+
"loss": 0.0921,
|
| 427 |
+
"step": 4500
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 3.8336807002917883,
|
| 431 |
+
"grad_norm": 1.5991489887237549,
|
| 432 |
+
"learning_rate": 5.463926354099764e-05,
|
| 433 |
+
"loss": 0.1044,
|
| 434 |
+
"step": 4600
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.9170487703209673,
|
| 438 |
+
"grad_norm": 1.71983003616333,
|
| 439 |
+
"learning_rate": 5.426461143224149e-05,
|
| 440 |
+
"loss": 0.0891,
|
| 441 |
+
"step": 4700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 4.0,
|
| 445 |
+
"grad_norm": 0.6225905418395996,
|
| 446 |
+
"learning_rate": 5.3889959323485334e-05,
|
| 447 |
+
"loss": 0.0935,
|
| 448 |
+
"step": 4800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.0,
|
| 452 |
+
"eval_administration_accuracy": 0.9401709401709402,
|
| 453 |
+
"eval_administration_f1": 0.9351367023709064,
|
| 454 |
+
"eval_corruption_accuracy": 0.9568480300187617,
|
| 455 |
+
"eval_corruption_f1": 0.9560031856416787,
|
| 456 |
+
"eval_democracy_accuracy": 0.9563268709610173,
|
| 457 |
+
"eval_democracy_f1": 0.9539147719955986,
|
| 458 |
+
"eval_development_accuracy": 0.9181780279341255,
|
| 459 |
+
"eval_development_f1": 0.9124572322058647,
|
| 460 |
+
"eval_economy_accuracy": 0.929435063581405,
|
| 461 |
+
"eval_economy_f1": 0.9284126365185199,
|
| 462 |
+
"eval_education_accuracy": 0.9672712111736502,
|
| 463 |
+
"eval_education_f1": 0.9657318355589747,
|
| 464 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 465 |
+
"eval_environment_f1": 0.9793120114242748,
|
| 466 |
+
"eval_instability_accuracy": 0.9337085678549093,
|
| 467 |
+
"eval_instability_f1": 0.9360007265571828,
|
| 468 |
+
"eval_leadership_accuracy": 0.8413591828225975,
|
| 469 |
+
"eval_leadership_f1": 0.8351481808317307,
|
| 470 |
+
"eval_loss": 0.22106842696666718,
|
| 471 |
+
"eval_overall_accuracy": 0.9382079077201028,
|
| 472 |
+
"eval_overall_f1": 0.9358746589615409,
|
| 473 |
+
"eval_race_accuracy": 0.9561184073379195,
|
| 474 |
+
"eval_race_f1": 0.9543792719379139,
|
| 475 |
+
"eval_religion_accuracy": 0.9570564936418595,
|
| 476 |
+
"eval_religion_f1": 0.9560165149767178,
|
| 477 |
+
"eval_runtime": 10.9937,
|
| 478 |
+
"eval_safety_accuracy": 0.9215134459036898,
|
| 479 |
+
"eval_safety_f1": 0.9179828375191263,
|
| 480 |
+
"eval_samples_per_second": 872.679,
|
| 481 |
+
"eval_steps_per_second": 54.577,
|
| 482 |
+
"step": 4800
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.083368070029179,
|
| 486 |
+
"grad_norm": 0.6198378801345825,
|
| 487 |
+
"learning_rate": 5.351530721472918e-05,
|
| 488 |
+
"loss": 0.0658,
|
| 489 |
+
"step": 4900
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.166736140058358,
|
| 493 |
+
"grad_norm": 1.2718008756637573,
|
| 494 |
+
"learning_rate": 5.3140655105973025e-05,
|
| 495 |
+
"loss": 0.0636,
|
| 496 |
+
"step": 5000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.250104210087536,
|
| 500 |
+
"grad_norm": 1.476962685585022,
|
| 501 |
+
"learning_rate": 5.2766002997216864e-05,
|
| 502 |
+
"loss": 0.0742,
|
| 503 |
+
"step": 5100
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 4.333472280116715,
|
| 507 |
+
"grad_norm": 1.1485812664031982,
|
| 508 |
+
"learning_rate": 5.239135088846071e-05,
|
| 509 |
+
"loss": 0.0693,
|
| 510 |
+
"step": 5200
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 4.416840350145894,
|
| 514 |
+
"grad_norm": 1.2767895460128784,
|
| 515 |
+
"learning_rate": 5.2016698779704556e-05,
|
| 516 |
+
"loss": 0.0722,
|
| 517 |
+
"step": 5300
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 4.500208420175073,
|
| 521 |
+
"grad_norm": 0.6865003108978271,
|
| 522 |
+
"learning_rate": 5.16420466709484e-05,
|
| 523 |
+
"loss": 0.0701,
|
| 524 |
+
"step": 5400
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.583576490204251,
|
| 528 |
+
"grad_norm": 0.6567270755767822,
|
| 529 |
+
"learning_rate": 5.127114108327981e-05,
|
| 530 |
+
"loss": 0.069,
|
| 531 |
+
"step": 5500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.66694456023343,
|
| 535 |
+
"grad_norm": 1.1187883615493774,
|
| 536 |
+
"learning_rate": 5.0896488974523654e-05,
|
| 537 |
+
"loss": 0.0716,
|
| 538 |
+
"step": 5600
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.750312630262609,
|
| 542 |
+
"grad_norm": 1.3242968320846558,
|
| 543 |
+
"learning_rate": 5.05218368657675e-05,
|
| 544 |
+
"loss": 0.0702,
|
| 545 |
+
"step": 5700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 4.833680700291788,
|
| 549 |
+
"grad_norm": 1.458292841911316,
|
| 550 |
+
"learning_rate": 5.0147184757011346e-05,
|
| 551 |
+
"loss": 0.07,
|
| 552 |
+
"step": 5800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 4.917048770320967,
|
| 556 |
+
"grad_norm": 0.7626857757568359,
|
| 557 |
+
"learning_rate": 4.977253264825519e-05,
|
| 558 |
+
"loss": 0.0733,
|
| 559 |
+
"step": 5900
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 5.0,
|
| 563 |
+
"grad_norm": 0.7239888310432434,
|
| 564 |
+
"learning_rate": 4.939788053949904e-05,
|
| 565 |
+
"loss": 0.0675,
|
| 566 |
+
"step": 6000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.0,
|
| 570 |
+
"eval_administration_accuracy": 0.9306858453199917,
|
| 571 |
+
"eval_administration_f1": 0.9286500450585223,
|
| 572 |
+
"eval_corruption_accuracy": 0.9606003752345216,
|
| 573 |
+
"eval_corruption_f1": 0.9583563972269709,
|
| 574 |
+
"eval_democracy_accuracy": 0.9416301855326246,
|
| 575 |
+
"eval_democracy_f1": 0.9449257041063754,
|
| 576 |
+
"eval_development_accuracy": 0.9198457369189077,
|
| 577 |
+
"eval_development_f1": 0.9154714581737773,
|
| 578 |
+
"eval_economy_accuracy": 0.9344381905357515,
|
| 579 |
+
"eval_economy_f1": 0.9306995601866777,
|
| 580 |
+
"eval_education_accuracy": 0.9673754429851991,
|
| 581 |
+
"eval_education_f1": 0.9648349073741065,
|
| 582 |
+
"eval_environment_accuracy": 0.9803001876172608,
|
| 583 |
+
"eval_environment_f1": 0.9792105415103222,
|
| 584 |
+
"eval_instability_accuracy": 0.944757139879091,
|
| 585 |
+
"eval_instability_f1": 0.9391731690067155,
|
| 586 |
+
"eval_leadership_accuracy": 0.8202001250781739,
|
| 587 |
+
"eval_leadership_f1": 0.8225964840638132,
|
| 588 |
+
"eval_loss": 0.2460281401872635,
|
| 589 |
+
"eval_overall_accuracy": 0.9360364116461678,
|
| 590 |
+
"eval_overall_f1": 0.9339890155373726,
|
| 591 |
+
"eval_race_accuracy": 0.9508025849489264,
|
| 592 |
+
"eval_race_f1": 0.9501627076296991,
|
| 593 |
+
"eval_religion_accuracy": 0.9597665207421305,
|
| 594 |
+
"eval_religion_f1": 0.9575416931106416,
|
| 595 |
+
"eval_runtime": 10.9751,
|
| 596 |
+
"eval_safety_accuracy": 0.9220346049614342,
|
| 597 |
+
"eval_safety_f1": 0.916245519000848,
|
| 598 |
+
"eval_samples_per_second": 874.157,
|
| 599 |
+
"eval_steps_per_second": 54.669,
|
| 600 |
+
"step": 6000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 5.083368070029179,
|
| 604 |
+
"grad_norm": 1.2741373777389526,
|
| 605 |
+
"learning_rate": 4.902322843074288e-05,
|
| 606 |
+
"loss": 0.0551,
|
| 607 |
+
"step": 6100
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 5.166736140058358,
|
| 611 |
+
"grad_norm": 0.5817021727561951,
|
| 612 |
+
"learning_rate": 4.864857632198673e-05,
|
| 613 |
+
"loss": 0.0505,
|
| 614 |
+
"step": 6200
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 5.250104210087536,
|
| 618 |
+
"grad_norm": 1.0689315795898438,
|
| 619 |
+
"learning_rate": 4.827392421323057e-05,
|
| 620 |
+
"loss": 0.049,
|
| 621 |
+
"step": 6300
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 5.333472280116715,
|
| 625 |
+
"grad_norm": 1.5673719644546509,
|
| 626 |
+
"learning_rate": 4.789927210447441e-05,
|
| 627 |
+
"loss": 0.0554,
|
| 628 |
+
"step": 6400
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 5.416840350145894,
|
| 632 |
+
"grad_norm": 1.1274124383926392,
|
| 633 |
+
"learning_rate": 4.752461999571826e-05,
|
| 634 |
+
"loss": 0.0523,
|
| 635 |
+
"step": 6500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 5.500208420175073,
|
| 639 |
+
"grad_norm": 0.836388349533081,
|
| 640 |
+
"learning_rate": 4.7149967886962105e-05,
|
| 641 |
+
"loss": 0.0516,
|
| 642 |
+
"step": 6600
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 5.583576490204251,
|
| 646 |
+
"grad_norm": 1.7012029886245728,
|
| 647 |
+
"learning_rate": 4.677531577820595e-05,
|
| 648 |
+
"loss": 0.0559,
|
| 649 |
+
"step": 6700
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 5.66694456023343,
|
| 653 |
+
"grad_norm": 1.0244345664978027,
|
| 654 |
+
"learning_rate": 4.6400663669449796e-05,
|
| 655 |
+
"loss": 0.0552,
|
| 656 |
+
"step": 6800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.750312630262609,
|
| 660 |
+
"grad_norm": 1.2528122663497925,
|
| 661 |
+
"learning_rate": 4.602601156069364e-05,
|
| 662 |
+
"loss": 0.0508,
|
| 663 |
+
"step": 6900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.833680700291788,
|
| 667 |
+
"grad_norm": 1.044662594795227,
|
| 668 |
+
"learning_rate": 4.565135945193749e-05,
|
| 669 |
+
"loss": 0.0531,
|
| 670 |
+
"step": 7000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 5.917048770320967,
|
| 674 |
+
"grad_norm": 0.8813854455947876,
|
| 675 |
+
"learning_rate": 4.5276707343181333e-05,
|
| 676 |
+
"loss": 0.0482,
|
| 677 |
+
"step": 7100
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 6.0,
|
| 681 |
+
"grad_norm": 1.4556822776794434,
|
| 682 |
+
"learning_rate": 4.490205523442518e-05,
|
| 683 |
+
"loss": 0.0516,
|
| 684 |
+
"step": 7200
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 6.0,
|
| 688 |
+
"eval_administration_accuracy": 0.9356889722743381,
|
| 689 |
+
"eval_administration_f1": 0.9317820245044639,
|
| 690 |
+
"eval_corruption_accuracy": 0.9579945799457995,
|
| 691 |
+
"eval_corruption_f1": 0.9568343721711066,
|
| 692 |
+
"eval_democracy_accuracy": 0.9589326662497394,
|
| 693 |
+
"eval_democracy_f1": 0.95607216308818,
|
| 694 |
+
"eval_development_accuracy": 0.9168230143839899,
|
| 695 |
+
"eval_development_f1": 0.9143401221807914,
|
| 696 |
+
"eval_economy_accuracy": 0.9308943089430894,
|
| 697 |
+
"eval_economy_f1": 0.9294738147048666,
|
| 698 |
+
"eval_education_accuracy": 0.9669585157390036,
|
| 699 |
+
"eval_education_f1": 0.965464450546169,
|
| 700 |
+
"eval_environment_accuracy": 0.9786324786324786,
|
| 701 |
+
"eval_environment_f1": 0.977778980162068,
|
| 702 |
+
"eval_instability_accuracy": 0.9441317490097978,
|
| 703 |
+
"eval_instability_f1": 0.9395705042517813,
|
| 704 |
+
"eval_leadership_accuracy": 0.8412549510110485,
|
| 705 |
+
"eval_leadership_f1": 0.8395230039371011,
|
| 706 |
+
"eval_loss": 0.26496145129203796,
|
| 707 |
+
"eval_overall_accuracy": 0.938372941421722,
|
| 708 |
+
"eval_overall_f1": 0.9362734744251768,
|
| 709 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 710 |
+
"eval_race_f1": 0.9519598766506976,
|
| 711 |
+
"eval_religion_accuracy": 0.9579945799457995,
|
| 712 |
+
"eval_religion_f1": 0.9568944771373694,
|
| 713 |
+
"eval_runtime": 10.8844,
|
| 714 |
+
"eval_safety_accuracy": 0.9170314780070877,
|
| 715 |
+
"eval_safety_f1": 0.915587903767528,
|
| 716 |
+
"eval_samples_per_second": 881.445,
|
| 717 |
+
"eval_steps_per_second": 55.125,
|
| 718 |
+
"step": 7200
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 6.083368070029179,
|
| 722 |
+
"grad_norm": 0.7286815643310547,
|
| 723 |
+
"learning_rate": 4.452740312566902e-05,
|
| 724 |
+
"loss": 0.0366,
|
| 725 |
+
"step": 7300
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 6.166736140058358,
|
| 729 |
+
"grad_norm": 2.7514851093292236,
|
| 730 |
+
"learning_rate": 4.4152751016912864e-05,
|
| 731 |
+
"loss": 0.0392,
|
| 732 |
+
"step": 7400
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 6.250104210087536,
|
| 736 |
+
"grad_norm": 1.0808284282684326,
|
| 737 |
+
"learning_rate": 4.377809890815671e-05,
|
| 738 |
+
"loss": 0.0365,
|
| 739 |
+
"step": 7500
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 6.333472280116715,
|
| 743 |
+
"grad_norm": 0.5677124261856079,
|
| 744 |
+
"learning_rate": 4.3403446799400555e-05,
|
| 745 |
+
"loss": 0.0439,
|
| 746 |
+
"step": 7600
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 6.416840350145894,
|
| 750 |
+
"grad_norm": 0.7869217991828918,
|
| 751 |
+
"learning_rate": 4.30287946906444e-05,
|
| 752 |
+
"loss": 0.034,
|
| 753 |
+
"step": 7700
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 6.500208420175073,
|
| 757 |
+
"grad_norm": 0.6197104454040527,
|
| 758 |
+
"learning_rate": 4.265414258188825e-05,
|
| 759 |
+
"loss": 0.0375,
|
| 760 |
+
"step": 7800
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 6.583576490204251,
|
| 764 |
+
"grad_norm": 1.126968264579773,
|
| 765 |
+
"learning_rate": 4.228323699421965e-05,
|
| 766 |
+
"loss": 0.0437,
|
| 767 |
+
"step": 7900
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 6.66694456023343,
|
| 771 |
+
"grad_norm": 0.6215279698371887,
|
| 772 |
+
"learning_rate": 4.190858488546349e-05,
|
| 773 |
+
"loss": 0.0413,
|
| 774 |
+
"step": 8000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 6.750312630262609,
|
| 778 |
+
"grad_norm": 1.0289812088012695,
|
| 779 |
+
"learning_rate": 4.153393277670734e-05,
|
| 780 |
+
"loss": 0.0383,
|
| 781 |
+
"step": 8100
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 6.833680700291788,
|
| 785 |
+
"grad_norm": 1.2618753910064697,
|
| 786 |
+
"learning_rate": 4.1159280667951184e-05,
|
| 787 |
+
"loss": 0.037,
|
| 788 |
+
"step": 8200
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 6.917048770320967,
|
| 792 |
+
"grad_norm": 0.48271915316581726,
|
| 793 |
+
"learning_rate": 4.078462855919502e-05,
|
| 794 |
+
"loss": 0.0403,
|
| 795 |
+
"step": 8300
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 7.0,
|
| 799 |
+
"grad_norm": 1.4230297803878784,
|
| 800 |
+
"learning_rate": 4.040997645043887e-05,
|
| 801 |
+
"loss": 0.0421,
|
| 802 |
+
"step": 8400
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 7.0,
|
| 806 |
+
"eval_administration_accuracy": 0.9262038774233896,
|
| 807 |
+
"eval_administration_f1": 0.9271433160043757,
|
| 808 |
+
"eval_corruption_accuracy": 0.9570564936418595,
|
| 809 |
+
"eval_corruption_f1": 0.9567097147624816,
|
| 810 |
+
"eval_democracy_accuracy": 0.9531999166145507,
|
| 811 |
+
"eval_democracy_f1": 0.9521445432226068,
|
| 812 |
+
"eval_development_accuracy": 0.9057744423598082,
|
| 813 |
+
"eval_development_f1": 0.9056708555901262,
|
| 814 |
+
"eval_economy_accuracy": 0.9282885136543673,
|
| 815 |
+
"eval_economy_f1": 0.926345436006994,
|
| 816 |
+
"eval_education_accuracy": 0.962059620596206,
|
| 817 |
+
"eval_education_f1": 0.961835890625152,
|
| 818 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 819 |
+
"eval_environment_f1": 0.9790877092918502,
|
| 820 |
+
"eval_instability_accuracy": 0.9382947675630603,
|
| 821 |
+
"eval_instability_f1": 0.9376909954205174,
|
| 822 |
+
"eval_leadership_accuracy": 0.8362518240567021,
|
| 823 |
+
"eval_leadership_f1": 0.8378707787677608,
|
| 824 |
+
"eval_loss": 0.2829027473926544,
|
| 825 |
+
"eval_overall_accuracy": 0.9348638037662428,
|
| 826 |
+
"eval_overall_f1": 0.9341947646002294,
|
| 827 |
+
"eval_race_accuracy": 0.9518449030644153,
|
| 828 |
+
"eval_race_f1": 0.9506769394120873,
|
| 829 |
+
"eval_religion_accuracy": 0.9585157390035439,
|
| 830 |
+
"eval_religion_f1": 0.9579338409216822,
|
| 831 |
+
"eval_runtime": 10.9473,
|
| 832 |
+
"eval_safety_accuracy": 0.9203668959766521,
|
| 833 |
+
"eval_safety_f1": 0.917227155177119,
|
| 834 |
+
"eval_samples_per_second": 876.38,
|
| 835 |
+
"eval_steps_per_second": 54.808,
|
| 836 |
+
"step": 8400
|
| 837 |
+
}
|
| 838 |
+
],
|
| 839 |
+
"logging_steps": 100,
|
| 840 |
+
"max_steps": 19184,
|
| 841 |
+
"num_input_tokens_seen": 0,
|
| 842 |
+
"num_train_epochs": 16,
|
| 843 |
+
"save_steps": 500,
|
| 844 |
+
"stateful_callbacks": {
|
| 845 |
+
"EarlyStoppingCallback": {
|
| 846 |
+
"args": {
|
| 847 |
+
"early_stopping_patience": 2,
|
| 848 |
+
"early_stopping_threshold": 0.0
|
| 849 |
+
},
|
| 850 |
+
"attributes": {
|
| 851 |
+
"early_stopping_patience_counter": 1
|
| 852 |
+
}
|
| 853 |
+
},
|
| 854 |
+
"TrainerControl": {
|
| 855 |
+
"args": {
|
| 856 |
+
"should_epoch_stop": false,
|
| 857 |
+
"should_evaluate": false,
|
| 858 |
+
"should_log": false,
|
| 859 |
+
"should_save": true,
|
| 860 |
+
"should_training_stop": false
|
| 861 |
+
},
|
| 862 |
+
"attributes": {}
|
| 863 |
+
}
|
| 864 |
+
},
|
| 865 |
+
"total_flos": 7.070187082265395e+16,
|
| 866 |
+
"train_batch_size": 16,
|
| 867 |
+
"trial_name": null,
|
| 868 |
+
"trial_params": {
|
| 869 |
+
"gradient_accumulation_steps": 2,
|
| 870 |
+
"learning_rate": 7e-05,
|
| 871 |
+
"num_train_epochs": 16
|
| 872 |
+
}
|
| 873 |
+
}
|
run-2/checkpoint-8400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a417688ba3a38960888d219af94bafd1d9551df9005996513f2ac72dfc76c145
|
| 3 |
+
size 5368
|
run-2/checkpoint-9600/config.json
ADDED
|
@@ -0,0 +1,125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3",
|
| 16 |
+
"4": "LABEL_4",
|
| 17 |
+
"5": "LABEL_5",
|
| 18 |
+
"6": "LABEL_6",
|
| 19 |
+
"7": "LABEL_7",
|
| 20 |
+
"8": "LABEL_8",
|
| 21 |
+
"9": "LABEL_9",
|
| 22 |
+
"10": "LABEL_10",
|
| 23 |
+
"11": "LABEL_11",
|
| 24 |
+
"12": "LABEL_12",
|
| 25 |
+
"13": "LABEL_13",
|
| 26 |
+
"14": "LABEL_14",
|
| 27 |
+
"15": "LABEL_15",
|
| 28 |
+
"16": "LABEL_16",
|
| 29 |
+
"17": "LABEL_17",
|
| 30 |
+
"18": "LABEL_18",
|
| 31 |
+
"19": "LABEL_19",
|
| 32 |
+
"20": "LABEL_20",
|
| 33 |
+
"21": "LABEL_21",
|
| 34 |
+
"22": "LABEL_22",
|
| 35 |
+
"23": "LABEL_23",
|
| 36 |
+
"24": "LABEL_24",
|
| 37 |
+
"25": "LABEL_25",
|
| 38 |
+
"26": "LABEL_26",
|
| 39 |
+
"27": "LABEL_27",
|
| 40 |
+
"28": "LABEL_28",
|
| 41 |
+
"29": "LABEL_29",
|
| 42 |
+
"30": "LABEL_30",
|
| 43 |
+
"31": "LABEL_31",
|
| 44 |
+
"32": "LABEL_32",
|
| 45 |
+
"33": "LABEL_33",
|
| 46 |
+
"34": "LABEL_34",
|
| 47 |
+
"35": "LABEL_35",
|
| 48 |
+
"36": "LABEL_36",
|
| 49 |
+
"37": "LABEL_37",
|
| 50 |
+
"38": "LABEL_38",
|
| 51 |
+
"39": "LABEL_39",
|
| 52 |
+
"40": "LABEL_40",
|
| 53 |
+
"41": "LABEL_41",
|
| 54 |
+
"42": "LABEL_42",
|
| 55 |
+
"43": "LABEL_43",
|
| 56 |
+
"44": "LABEL_44",
|
| 57 |
+
"45": "LABEL_45",
|
| 58 |
+
"46": "LABEL_46",
|
| 59 |
+
"47": "LABEL_47"
|
| 60 |
+
},
|
| 61 |
+
"initializer_range": 0.02,
|
| 62 |
+
"intermediate_size": 3072,
|
| 63 |
+
"label2id": {
|
| 64 |
+
"LABEL_0": 0,
|
| 65 |
+
"LABEL_1": 1,
|
| 66 |
+
"LABEL_10": 10,
|
| 67 |
+
"LABEL_11": 11,
|
| 68 |
+
"LABEL_12": 12,
|
| 69 |
+
"LABEL_13": 13,
|
| 70 |
+
"LABEL_14": 14,
|
| 71 |
+
"LABEL_15": 15,
|
| 72 |
+
"LABEL_16": 16,
|
| 73 |
+
"LABEL_17": 17,
|
| 74 |
+
"LABEL_18": 18,
|
| 75 |
+
"LABEL_19": 19,
|
| 76 |
+
"LABEL_2": 2,
|
| 77 |
+
"LABEL_20": 20,
|
| 78 |
+
"LABEL_21": 21,
|
| 79 |
+
"LABEL_22": 22,
|
| 80 |
+
"LABEL_23": 23,
|
| 81 |
+
"LABEL_24": 24,
|
| 82 |
+
"LABEL_25": 25,
|
| 83 |
+
"LABEL_26": 26,
|
| 84 |
+
"LABEL_27": 27,
|
| 85 |
+
"LABEL_28": 28,
|
| 86 |
+
"LABEL_29": 29,
|
| 87 |
+
"LABEL_3": 3,
|
| 88 |
+
"LABEL_30": 30,
|
| 89 |
+
"LABEL_31": 31,
|
| 90 |
+
"LABEL_32": 32,
|
| 91 |
+
"LABEL_33": 33,
|
| 92 |
+
"LABEL_34": 34,
|
| 93 |
+
"LABEL_35": 35,
|
| 94 |
+
"LABEL_36": 36,
|
| 95 |
+
"LABEL_37": 37,
|
| 96 |
+
"LABEL_38": 38,
|
| 97 |
+
"LABEL_39": 39,
|
| 98 |
+
"LABEL_4": 4,
|
| 99 |
+
"LABEL_40": 40,
|
| 100 |
+
"LABEL_41": 41,
|
| 101 |
+
"LABEL_42": 42,
|
| 102 |
+
"LABEL_43": 43,
|
| 103 |
+
"LABEL_44": 44,
|
| 104 |
+
"LABEL_45": 45,
|
| 105 |
+
"LABEL_46": 46,
|
| 106 |
+
"LABEL_47": 47,
|
| 107 |
+
"LABEL_5": 5,
|
| 108 |
+
"LABEL_6": 6,
|
| 109 |
+
"LABEL_7": 7,
|
| 110 |
+
"LABEL_8": 8,
|
| 111 |
+
"LABEL_9": 9
|
| 112 |
+
},
|
| 113 |
+
"layer_norm_eps": 1e-12,
|
| 114 |
+
"max_position_embeddings": 512,
|
| 115 |
+
"model_type": "bert",
|
| 116 |
+
"num_attention_heads": 12,
|
| 117 |
+
"num_hidden_layers": 12,
|
| 118 |
+
"pad_token_id": 0,
|
| 119 |
+
"position_embedding_type": "absolute",
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.50.3",
|
| 122 |
+
"type_vocab_size": 2,
|
| 123 |
+
"use_cache": true,
|
| 124 |
+
"vocab_size": 30522
|
| 125 |
+
}
|
run-2/checkpoint-9600/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c340072dfef17fac21351cdcf69e3e3cfd8e691bf5f156e56c6e22e7dd9b9ffb
|
| 3 |
+
size 438100144
|
run-2/checkpoint-9600/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a39aa8c3124c59be8892b8efe52e85a917a20955115486b15ce5c1cd510cb479
|
| 3 |
+
size 876321402
|
run-2/checkpoint-9600/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ffad3d9471fb61b3f42fea2e7138fc106169adb362290f6561f26126680d7263
|
| 3 |
+
size 14244
|
run-2/checkpoint-9600/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2ae9ad4181ffc3e9df875e777e97d2f9b7c255e21df065b6238ff529b264c44
|
| 3 |
+
size 988
|
run-2/checkpoint-9600/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71336da14085796816d87de8a9778757f36434ba98e37214737395052ada44cd
|
| 3 |
+
size 1064
|
run-2/checkpoint-9600/trainer_state.json
ADDED
|
@@ -0,0 +1,991 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7200,
|
| 3 |
+
"best_metric": 0.9362734744251768,
|
| 4 |
+
"best_model_checkpoint": "./results/run-2/checkpoint-7200",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9600,
|
| 8 |
+
"is_hyper_param_search": true,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.08336807002917883,
|
| 14 |
+
"grad_norm": 1.2878578901290894,
|
| 15 |
+
"learning_rate": 1.4e-05,
|
| 16 |
+
"loss": 1.2644,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.16673614005835766,
|
| 21 |
+
"grad_norm": 1.0009047985076904,
|
| 22 |
+
"learning_rate": 2.8e-05,
|
| 23 |
+
"loss": 0.4758,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.25010421008753647,
|
| 28 |
+
"grad_norm": 0.7635347843170166,
|
| 29 |
+
"learning_rate": 4.2e-05,
|
| 30 |
+
"loss": 0.4054,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.3334722801167153,
|
| 35 |
+
"grad_norm": 0.8168604373931885,
|
| 36 |
+
"learning_rate": 5.6e-05,
|
| 37 |
+
"loss": 0.3468,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4168403501458941,
|
| 42 |
+
"grad_norm": 1.1069319248199463,
|
| 43 |
+
"learning_rate": 7e-05,
|
| 44 |
+
"loss": 0.3051,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5002084201750729,
|
| 49 |
+
"grad_norm": 1.015086054801941,
|
| 50 |
+
"learning_rate": 6.962534789124383e-05,
|
| 51 |
+
"loss": 0.3044,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.5835764902042517,
|
| 56 |
+
"grad_norm": 0.8430781364440918,
|
| 57 |
+
"learning_rate": 6.925069578248769e-05,
|
| 58 |
+
"loss": 0.2742,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.6669445602334306,
|
| 63 |
+
"grad_norm": 0.950409471988678,
|
| 64 |
+
"learning_rate": 6.887604367373152e-05,
|
| 65 |
+
"loss": 0.2573,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.7503126302626094,
|
| 70 |
+
"grad_norm": 1.171885371208191,
|
| 71 |
+
"learning_rate": 6.850139156497538e-05,
|
| 72 |
+
"loss": 0.2517,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.8336807002917882,
|
| 77 |
+
"grad_norm": 0.9133914709091187,
|
| 78 |
+
"learning_rate": 6.812673945621922e-05,
|
| 79 |
+
"loss": 0.2435,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.9170487703209671,
|
| 84 |
+
"grad_norm": 0.7939682602882385,
|
| 85 |
+
"learning_rate": 6.775208734746307e-05,
|
| 86 |
+
"loss": 0.2373,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.0,
|
| 91 |
+
"grad_norm": 0.49670639634132385,
|
| 92 |
+
"learning_rate": 6.737743523870691e-05,
|
| 93 |
+
"loss": 0.2363,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.0,
|
| 98 |
+
"eval_administration_accuracy": 0.9302689180737961,
|
| 99 |
+
"eval_administration_f1": 0.9077189587684091,
|
| 100 |
+
"eval_corruption_accuracy": 0.9534083802376485,
|
| 101 |
+
"eval_corruption_f1": 0.9474431988592255,
|
| 102 |
+
"eval_democracy_accuracy": 0.9473629351678132,
|
| 103 |
+
"eval_democracy_f1": 0.9358042521135516,
|
| 104 |
+
"eval_development_accuracy": 0.9094225557640192,
|
| 105 |
+
"eval_development_f1": 0.8937523747915405,
|
| 106 |
+
"eval_economy_accuracy": 0.924119241192412,
|
| 107 |
+
"eval_economy_f1": 0.9156575066932242,
|
| 108 |
+
"eval_education_accuracy": 0.9621638524077548,
|
| 109 |
+
"eval_education_f1": 0.9583809787976314,
|
| 110 |
+
"eval_environment_accuracy": 0.9775901605169898,
|
| 111 |
+
"eval_environment_f1": 0.9754739686963207,
|
| 112 |
+
"eval_instability_accuracy": 0.9422555764019178,
|
| 113 |
+
"eval_instability_f1": 0.9343996801103537,
|
| 114 |
+
"eval_leadership_accuracy": 0.7970606629143214,
|
| 115 |
+
"eval_leadership_f1": 0.7702375495381067,
|
| 116 |
+
"eval_loss": 0.22745274007320404,
|
| 117 |
+
"eval_overall_accuracy": 0.9296261552359112,
|
| 118 |
+
"eval_overall_f1": 0.9199542168750479,
|
| 119 |
+
"eval_race_accuracy": 0.9505941213258287,
|
| 120 |
+
"eval_race_f1": 0.9483610849847235,
|
| 121 |
+
"eval_religion_accuracy": 0.9505941213258287,
|
| 122 |
+
"eval_religion_f1": 0.9489643730683611,
|
| 123 |
+
"eval_runtime": 11.0331,
|
| 124 |
+
"eval_safety_accuracy": 0.9106733375026058,
|
| 125 |
+
"eval_safety_f1": 0.9032566760791263,
|
| 126 |
+
"eval_samples_per_second": 869.566,
|
| 127 |
+
"eval_steps_per_second": 54.382,
|
| 128 |
+
"step": 1200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.0833680700291788,
|
| 132 |
+
"grad_norm": 0.949184238910675,
|
| 133 |
+
"learning_rate": 6.700278312995076e-05,
|
| 134 |
+
"loss": 0.1968,
|
| 135 |
+
"step": 1300
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.1667361400583576,
|
| 139 |
+
"grad_norm": 0.7735128402709961,
|
| 140 |
+
"learning_rate": 6.66281310211946e-05,
|
| 141 |
+
"loss": 0.1924,
|
| 142 |
+
"step": 1400
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.2501042100875366,
|
| 146 |
+
"grad_norm": 1.0187016725540161,
|
| 147 |
+
"learning_rate": 6.625347891243845e-05,
|
| 148 |
+
"loss": 0.1945,
|
| 149 |
+
"step": 1500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 1.3334722801167154,
|
| 153 |
+
"grad_norm": 0.9707878828048706,
|
| 154 |
+
"learning_rate": 6.587882680368229e-05,
|
| 155 |
+
"loss": 0.1931,
|
| 156 |
+
"step": 1600
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 1.4168403501458942,
|
| 160 |
+
"grad_norm": 1.4602934122085571,
|
| 161 |
+
"learning_rate": 6.550417469492613e-05,
|
| 162 |
+
"loss": 0.1967,
|
| 163 |
+
"step": 1700
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.500208420175073,
|
| 167 |
+
"grad_norm": 1.2449183464050293,
|
| 168 |
+
"learning_rate": 6.512952258616998e-05,
|
| 169 |
+
"loss": 0.1843,
|
| 170 |
+
"step": 1800
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.5835764902042517,
|
| 174 |
+
"grad_norm": 1.4943019151687622,
|
| 175 |
+
"learning_rate": 6.475487047741382e-05,
|
| 176 |
+
"loss": 0.1844,
|
| 177 |
+
"step": 1900
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.6669445602334307,
|
| 181 |
+
"grad_norm": 1.8496220111846924,
|
| 182 |
+
"learning_rate": 6.438021836865767e-05,
|
| 183 |
+
"loss": 0.1797,
|
| 184 |
+
"step": 2000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.7503126302626093,
|
| 188 |
+
"grad_norm": 1.0578210353851318,
|
| 189 |
+
"learning_rate": 6.400556625990151e-05,
|
| 190 |
+
"loss": 0.1875,
|
| 191 |
+
"step": 2100
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.8336807002917883,
|
| 195 |
+
"grad_norm": 1.0538480281829834,
|
| 196 |
+
"learning_rate": 6.363091415114536e-05,
|
| 197 |
+
"loss": 0.1685,
|
| 198 |
+
"step": 2200
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.917048770320967,
|
| 202 |
+
"grad_norm": 1.5351332426071167,
|
| 203 |
+
"learning_rate": 6.32562620423892e-05,
|
| 204 |
+
"loss": 0.1744,
|
| 205 |
+
"step": 2300
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.0,
|
| 209 |
+
"grad_norm": 0.9578828811645508,
|
| 210 |
+
"learning_rate": 6.288160993363306e-05,
|
| 211 |
+
"loss": 0.1811,
|
| 212 |
+
"step": 2400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.0,
|
| 216 |
+
"eval_administration_accuracy": 0.9344381905357515,
|
| 217 |
+
"eval_administration_f1": 0.9296143144580357,
|
| 218 |
+
"eval_corruption_accuracy": 0.9604961434229727,
|
| 219 |
+
"eval_corruption_f1": 0.9577803781696625,
|
| 220 |
+
"eval_democracy_accuracy": 0.9534083802376485,
|
| 221 |
+
"eval_democracy_f1": 0.9434755362160818,
|
| 222 |
+
"eval_development_accuracy": 0.9193245778611632,
|
| 223 |
+
"eval_development_f1": 0.9132597892423039,
|
| 224 |
+
"eval_economy_accuracy": 0.931311236189285,
|
| 225 |
+
"eval_economy_f1": 0.92884213698996,
|
| 226 |
+
"eval_education_accuracy": 0.9653950385657702,
|
| 227 |
+
"eval_education_f1": 0.963862144635937,
|
| 228 |
+
"eval_environment_accuracy": 0.979153637690223,
|
| 229 |
+
"eval_environment_f1": 0.9783610154594226,
|
| 230 |
+
"eval_instability_accuracy": 0.9459036898061288,
|
| 231 |
+
"eval_instability_f1": 0.9423381513536221,
|
| 232 |
+
"eval_leadership_accuracy": 0.8387533875338753,
|
| 233 |
+
"eval_leadership_f1": 0.8323519207670947,
|
| 234 |
+
"eval_loss": 0.2031705528497696,
|
| 235 |
+
"eval_overall_accuracy": 0.9372177055103883,
|
| 236 |
+
"eval_overall_f1": 0.9339233416905627,
|
| 237 |
+
"eval_race_accuracy": 0.9516364394413175,
|
| 238 |
+
"eval_race_f1": 0.9498984068182696,
|
| 239 |
+
"eval_religion_accuracy": 0.9558057119032729,
|
| 240 |
+
"eval_religion_f1": 0.9554428854878803,
|
| 241 |
+
"eval_runtime": 11.0378,
|
| 242 |
+
"eval_safety_accuracy": 0.9109860329372524,
|
| 243 |
+
"eval_safety_f1": 0.9118534206884823,
|
| 244 |
+
"eval_samples_per_second": 869.197,
|
| 245 |
+
"eval_steps_per_second": 54.359,
|
| 246 |
+
"step": 2400
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 2.083368070029179,
|
| 250 |
+
"grad_norm": 1.0708891153335571,
|
| 251 |
+
"learning_rate": 6.25069578248769e-05,
|
| 252 |
+
"loss": 0.1338,
|
| 253 |
+
"step": 2500
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 2.1667361400583576,
|
| 257 |
+
"grad_norm": 1.0800749063491821,
|
| 258 |
+
"learning_rate": 6.213230571612073e-05,
|
| 259 |
+
"loss": 0.1268,
|
| 260 |
+
"step": 2600
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 2.2501042100875366,
|
| 264 |
+
"grad_norm": 1.3577998876571655,
|
| 265 |
+
"learning_rate": 6.175765360736459e-05,
|
| 266 |
+
"loss": 0.1226,
|
| 267 |
+
"step": 2700
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 2.333472280116715,
|
| 271 |
+
"grad_norm": 1.325534701347351,
|
| 272 |
+
"learning_rate": 6.138300149860843e-05,
|
| 273 |
+
"loss": 0.1258,
|
| 274 |
+
"step": 2800
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 2.416840350145894,
|
| 278 |
+
"grad_norm": 1.7846063375473022,
|
| 279 |
+
"learning_rate": 6.100834938985228e-05,
|
| 280 |
+
"loss": 0.1308,
|
| 281 |
+
"step": 2900
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.500208420175073,
|
| 285 |
+
"grad_norm": 1.61028254032135,
|
| 286 |
+
"learning_rate": 6.0633697281096124e-05,
|
| 287 |
+
"loss": 0.1305,
|
| 288 |
+
"step": 3000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.5835764902042517,
|
| 292 |
+
"grad_norm": 1.2228693962097168,
|
| 293 |
+
"learning_rate": 6.025904517233996e-05,
|
| 294 |
+
"loss": 0.1315,
|
| 295 |
+
"step": 3100
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 2.6669445602334307,
|
| 299 |
+
"grad_norm": 0.9486532807350159,
|
| 300 |
+
"learning_rate": 5.988439306358381e-05,
|
| 301 |
+
"loss": 0.1337,
|
| 302 |
+
"step": 3200
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 2.7503126302626093,
|
| 306 |
+
"grad_norm": 1.3199305534362793,
|
| 307 |
+
"learning_rate": 5.9509740954827654e-05,
|
| 308 |
+
"loss": 0.132,
|
| 309 |
+
"step": 3300
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 2.8336807002917883,
|
| 313 |
+
"grad_norm": 1.8821613788604736,
|
| 314 |
+
"learning_rate": 5.91350888460715e-05,
|
| 315 |
+
"loss": 0.1258,
|
| 316 |
+
"step": 3400
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 2.9170487703209673,
|
| 320 |
+
"grad_norm": 1.2816129922866821,
|
| 321 |
+
"learning_rate": 5.8760436737315346e-05,
|
| 322 |
+
"loss": 0.1282,
|
| 323 |
+
"step": 3500
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 3.0,
|
| 327 |
+
"grad_norm": 5.737521171569824,
|
| 328 |
+
"learning_rate": 5.838578462855919e-05,
|
| 329 |
+
"loss": 0.1297,
|
| 330 |
+
"step": 3600
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 3.0,
|
| 334 |
+
"eval_administration_accuracy": 0.9375651448822181,
|
| 335 |
+
"eval_administration_f1": 0.9309689641816119,
|
| 336 |
+
"eval_corruption_accuracy": 0.9451740671252866,
|
| 337 |
+
"eval_corruption_f1": 0.9476001634812502,
|
| 338 |
+
"eval_democracy_accuracy": 0.9508025849489264,
|
| 339 |
+
"eval_democracy_f1": 0.9508783301626133,
|
| 340 |
+
"eval_development_accuracy": 0.921304982280592,
|
| 341 |
+
"eval_development_f1": 0.9145210502797096,
|
| 342 |
+
"eval_economy_accuracy": 0.930373149885345,
|
| 343 |
+
"eval_economy_f1": 0.9257602899924291,
|
| 344 |
+
"eval_education_accuracy": 0.9670627475505524,
|
| 345 |
+
"eval_education_f1": 0.9651765113114141,
|
| 346 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 347 |
+
"eval_environment_f1": 0.9785021531425564,
|
| 348 |
+
"eval_instability_accuracy": 0.9419428809672712,
|
| 349 |
+
"eval_instability_f1": 0.9412425903038227,
|
| 350 |
+
"eval_leadership_accuracy": 0.8284344381905357,
|
| 351 |
+
"eval_leadership_f1": 0.8287745355631846,
|
| 352 |
+
"eval_loss": 0.21378232538700104,
|
| 353 |
+
"eval_overall_accuracy": 0.9360450976304634,
|
| 354 |
+
"eval_overall_f1": 0.9337781089602167,
|
| 355 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 356 |
+
"eval_race_f1": 0.9510004306979282,
|
| 357 |
+
"eval_religion_accuracy": 0.955701480091724,
|
| 358 |
+
"eval_religion_f1": 0.9539633431565642,
|
| 359 |
+
"eval_runtime": 11.1303,
|
| 360 |
+
"eval_safety_accuracy": 0.919533041484261,
|
| 361 |
+
"eval_safety_f1": 0.9169489452495161,
|
| 362 |
+
"eval_samples_per_second": 861.972,
|
| 363 |
+
"eval_steps_per_second": 53.907,
|
| 364 |
+
"step": 3600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 3.083368070029179,
|
| 368 |
+
"grad_norm": 0.7353236079216003,
|
| 369 |
+
"learning_rate": 5.801113251980304e-05,
|
| 370 |
+
"loss": 0.0912,
|
| 371 |
+
"step": 3700
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 3.1667361400583576,
|
| 375 |
+
"grad_norm": 1.2753976583480835,
|
| 376 |
+
"learning_rate": 5.763648041104688e-05,
|
| 377 |
+
"loss": 0.0954,
|
| 378 |
+
"step": 3800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 3.2501042100875366,
|
| 382 |
+
"grad_norm": 1.4709268808364868,
|
| 383 |
+
"learning_rate": 5.726182830229073e-05,
|
| 384 |
+
"loss": 0.0936,
|
| 385 |
+
"step": 3900
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 3.333472280116715,
|
| 389 |
+
"grad_norm": 0.8533086180686951,
|
| 390 |
+
"learning_rate": 5.6887176193534574e-05,
|
| 391 |
+
"loss": 0.0927,
|
| 392 |
+
"step": 4000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 3.416840350145894,
|
| 396 |
+
"grad_norm": 1.1082295179367065,
|
| 397 |
+
"learning_rate": 5.6512524084778413e-05,
|
| 398 |
+
"loss": 0.091,
|
| 399 |
+
"step": 4100
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 3.500208420175073,
|
| 403 |
+
"grad_norm": 1.3943876028060913,
|
| 404 |
+
"learning_rate": 5.613787197602226e-05,
|
| 405 |
+
"loss": 0.0966,
|
| 406 |
+
"step": 4200
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.5835764902042517,
|
| 410 |
+
"grad_norm": 1.483786940574646,
|
| 411 |
+
"learning_rate": 5.5763219867266105e-05,
|
| 412 |
+
"loss": 0.0965,
|
| 413 |
+
"step": 4300
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 3.6669445602334307,
|
| 417 |
+
"grad_norm": 1.0739402770996094,
|
| 418 |
+
"learning_rate": 5.538856775850995e-05,
|
| 419 |
+
"loss": 0.0988,
|
| 420 |
+
"step": 4400
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 3.7503126302626093,
|
| 424 |
+
"grad_norm": 1.2540950775146484,
|
| 425 |
+
"learning_rate": 5.5013915649753796e-05,
|
| 426 |
+
"loss": 0.0921,
|
| 427 |
+
"step": 4500
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 3.8336807002917883,
|
| 431 |
+
"grad_norm": 1.5991489887237549,
|
| 432 |
+
"learning_rate": 5.463926354099764e-05,
|
| 433 |
+
"loss": 0.1044,
|
| 434 |
+
"step": 4600
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 3.9170487703209673,
|
| 438 |
+
"grad_norm": 1.71983003616333,
|
| 439 |
+
"learning_rate": 5.426461143224149e-05,
|
| 440 |
+
"loss": 0.0891,
|
| 441 |
+
"step": 4700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 4.0,
|
| 445 |
+
"grad_norm": 0.6225905418395996,
|
| 446 |
+
"learning_rate": 5.3889959323485334e-05,
|
| 447 |
+
"loss": 0.0935,
|
| 448 |
+
"step": 4800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.0,
|
| 452 |
+
"eval_administration_accuracy": 0.9401709401709402,
|
| 453 |
+
"eval_administration_f1": 0.9351367023709064,
|
| 454 |
+
"eval_corruption_accuracy": 0.9568480300187617,
|
| 455 |
+
"eval_corruption_f1": 0.9560031856416787,
|
| 456 |
+
"eval_democracy_accuracy": 0.9563268709610173,
|
| 457 |
+
"eval_democracy_f1": 0.9539147719955986,
|
| 458 |
+
"eval_development_accuracy": 0.9181780279341255,
|
| 459 |
+
"eval_development_f1": 0.9124572322058647,
|
| 460 |
+
"eval_economy_accuracy": 0.929435063581405,
|
| 461 |
+
"eval_economy_f1": 0.9284126365185199,
|
| 462 |
+
"eval_education_accuracy": 0.9672712111736502,
|
| 463 |
+
"eval_education_f1": 0.9657318355589747,
|
| 464 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 465 |
+
"eval_environment_f1": 0.9793120114242748,
|
| 466 |
+
"eval_instability_accuracy": 0.9337085678549093,
|
| 467 |
+
"eval_instability_f1": 0.9360007265571828,
|
| 468 |
+
"eval_leadership_accuracy": 0.8413591828225975,
|
| 469 |
+
"eval_leadership_f1": 0.8351481808317307,
|
| 470 |
+
"eval_loss": 0.22106842696666718,
|
| 471 |
+
"eval_overall_accuracy": 0.9382079077201028,
|
| 472 |
+
"eval_overall_f1": 0.9358746589615409,
|
| 473 |
+
"eval_race_accuracy": 0.9561184073379195,
|
| 474 |
+
"eval_race_f1": 0.9543792719379139,
|
| 475 |
+
"eval_religion_accuracy": 0.9570564936418595,
|
| 476 |
+
"eval_religion_f1": 0.9560165149767178,
|
| 477 |
+
"eval_runtime": 10.9937,
|
| 478 |
+
"eval_safety_accuracy": 0.9215134459036898,
|
| 479 |
+
"eval_safety_f1": 0.9179828375191263,
|
| 480 |
+
"eval_samples_per_second": 872.679,
|
| 481 |
+
"eval_steps_per_second": 54.577,
|
| 482 |
+
"step": 4800
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.083368070029179,
|
| 486 |
+
"grad_norm": 0.6198378801345825,
|
| 487 |
+
"learning_rate": 5.351530721472918e-05,
|
| 488 |
+
"loss": 0.0658,
|
| 489 |
+
"step": 4900
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.166736140058358,
|
| 493 |
+
"grad_norm": 1.2718008756637573,
|
| 494 |
+
"learning_rate": 5.3140655105973025e-05,
|
| 495 |
+
"loss": 0.0636,
|
| 496 |
+
"step": 5000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.250104210087536,
|
| 500 |
+
"grad_norm": 1.476962685585022,
|
| 501 |
+
"learning_rate": 5.2766002997216864e-05,
|
| 502 |
+
"loss": 0.0742,
|
| 503 |
+
"step": 5100
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 4.333472280116715,
|
| 507 |
+
"grad_norm": 1.1485812664031982,
|
| 508 |
+
"learning_rate": 5.239135088846071e-05,
|
| 509 |
+
"loss": 0.0693,
|
| 510 |
+
"step": 5200
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 4.416840350145894,
|
| 514 |
+
"grad_norm": 1.2767895460128784,
|
| 515 |
+
"learning_rate": 5.2016698779704556e-05,
|
| 516 |
+
"loss": 0.0722,
|
| 517 |
+
"step": 5300
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 4.500208420175073,
|
| 521 |
+
"grad_norm": 0.6865003108978271,
|
| 522 |
+
"learning_rate": 5.16420466709484e-05,
|
| 523 |
+
"loss": 0.0701,
|
| 524 |
+
"step": 5400
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 4.583576490204251,
|
| 528 |
+
"grad_norm": 0.6567270755767822,
|
| 529 |
+
"learning_rate": 5.127114108327981e-05,
|
| 530 |
+
"loss": 0.069,
|
| 531 |
+
"step": 5500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 4.66694456023343,
|
| 535 |
+
"grad_norm": 1.1187883615493774,
|
| 536 |
+
"learning_rate": 5.0896488974523654e-05,
|
| 537 |
+
"loss": 0.0716,
|
| 538 |
+
"step": 5600
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.750312630262609,
|
| 542 |
+
"grad_norm": 1.3242968320846558,
|
| 543 |
+
"learning_rate": 5.05218368657675e-05,
|
| 544 |
+
"loss": 0.0702,
|
| 545 |
+
"step": 5700
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 4.833680700291788,
|
| 549 |
+
"grad_norm": 1.458292841911316,
|
| 550 |
+
"learning_rate": 5.0147184757011346e-05,
|
| 551 |
+
"loss": 0.07,
|
| 552 |
+
"step": 5800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 4.917048770320967,
|
| 556 |
+
"grad_norm": 0.7626857757568359,
|
| 557 |
+
"learning_rate": 4.977253264825519e-05,
|
| 558 |
+
"loss": 0.0733,
|
| 559 |
+
"step": 5900
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 5.0,
|
| 563 |
+
"grad_norm": 0.7239888310432434,
|
| 564 |
+
"learning_rate": 4.939788053949904e-05,
|
| 565 |
+
"loss": 0.0675,
|
| 566 |
+
"step": 6000
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 5.0,
|
| 570 |
+
"eval_administration_accuracy": 0.9306858453199917,
|
| 571 |
+
"eval_administration_f1": 0.9286500450585223,
|
| 572 |
+
"eval_corruption_accuracy": 0.9606003752345216,
|
| 573 |
+
"eval_corruption_f1": 0.9583563972269709,
|
| 574 |
+
"eval_democracy_accuracy": 0.9416301855326246,
|
| 575 |
+
"eval_democracy_f1": 0.9449257041063754,
|
| 576 |
+
"eval_development_accuracy": 0.9198457369189077,
|
| 577 |
+
"eval_development_f1": 0.9154714581737773,
|
| 578 |
+
"eval_economy_accuracy": 0.9344381905357515,
|
| 579 |
+
"eval_economy_f1": 0.9306995601866777,
|
| 580 |
+
"eval_education_accuracy": 0.9673754429851991,
|
| 581 |
+
"eval_education_f1": 0.9648349073741065,
|
| 582 |
+
"eval_environment_accuracy": 0.9803001876172608,
|
| 583 |
+
"eval_environment_f1": 0.9792105415103222,
|
| 584 |
+
"eval_instability_accuracy": 0.944757139879091,
|
| 585 |
+
"eval_instability_f1": 0.9391731690067155,
|
| 586 |
+
"eval_leadership_accuracy": 0.8202001250781739,
|
| 587 |
+
"eval_leadership_f1": 0.8225964840638132,
|
| 588 |
+
"eval_loss": 0.2460281401872635,
|
| 589 |
+
"eval_overall_accuracy": 0.9360364116461678,
|
| 590 |
+
"eval_overall_f1": 0.9339890155373726,
|
| 591 |
+
"eval_race_accuracy": 0.9508025849489264,
|
| 592 |
+
"eval_race_f1": 0.9501627076296991,
|
| 593 |
+
"eval_religion_accuracy": 0.9597665207421305,
|
| 594 |
+
"eval_religion_f1": 0.9575416931106416,
|
| 595 |
+
"eval_runtime": 10.9751,
|
| 596 |
+
"eval_safety_accuracy": 0.9220346049614342,
|
| 597 |
+
"eval_safety_f1": 0.916245519000848,
|
| 598 |
+
"eval_samples_per_second": 874.157,
|
| 599 |
+
"eval_steps_per_second": 54.669,
|
| 600 |
+
"step": 6000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 5.083368070029179,
|
| 604 |
+
"grad_norm": 1.2741373777389526,
|
| 605 |
+
"learning_rate": 4.902322843074288e-05,
|
| 606 |
+
"loss": 0.0551,
|
| 607 |
+
"step": 6100
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 5.166736140058358,
|
| 611 |
+
"grad_norm": 0.5817021727561951,
|
| 612 |
+
"learning_rate": 4.864857632198673e-05,
|
| 613 |
+
"loss": 0.0505,
|
| 614 |
+
"step": 6200
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 5.250104210087536,
|
| 618 |
+
"grad_norm": 1.0689315795898438,
|
| 619 |
+
"learning_rate": 4.827392421323057e-05,
|
| 620 |
+
"loss": 0.049,
|
| 621 |
+
"step": 6300
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 5.333472280116715,
|
| 625 |
+
"grad_norm": 1.5673719644546509,
|
| 626 |
+
"learning_rate": 4.789927210447441e-05,
|
| 627 |
+
"loss": 0.0554,
|
| 628 |
+
"step": 6400
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 5.416840350145894,
|
| 632 |
+
"grad_norm": 1.1274124383926392,
|
| 633 |
+
"learning_rate": 4.752461999571826e-05,
|
| 634 |
+
"loss": 0.0523,
|
| 635 |
+
"step": 6500
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 5.500208420175073,
|
| 639 |
+
"grad_norm": 0.836388349533081,
|
| 640 |
+
"learning_rate": 4.7149967886962105e-05,
|
| 641 |
+
"loss": 0.0516,
|
| 642 |
+
"step": 6600
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 5.583576490204251,
|
| 646 |
+
"grad_norm": 1.7012029886245728,
|
| 647 |
+
"learning_rate": 4.677531577820595e-05,
|
| 648 |
+
"loss": 0.0559,
|
| 649 |
+
"step": 6700
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 5.66694456023343,
|
| 653 |
+
"grad_norm": 1.0244345664978027,
|
| 654 |
+
"learning_rate": 4.6400663669449796e-05,
|
| 655 |
+
"loss": 0.0552,
|
| 656 |
+
"step": 6800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.750312630262609,
|
| 660 |
+
"grad_norm": 1.2528122663497925,
|
| 661 |
+
"learning_rate": 4.602601156069364e-05,
|
| 662 |
+
"loss": 0.0508,
|
| 663 |
+
"step": 6900
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 5.833680700291788,
|
| 667 |
+
"grad_norm": 1.044662594795227,
|
| 668 |
+
"learning_rate": 4.565135945193749e-05,
|
| 669 |
+
"loss": 0.0531,
|
| 670 |
+
"step": 7000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 5.917048770320967,
|
| 674 |
+
"grad_norm": 0.8813854455947876,
|
| 675 |
+
"learning_rate": 4.5276707343181333e-05,
|
| 676 |
+
"loss": 0.0482,
|
| 677 |
+
"step": 7100
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 6.0,
|
| 681 |
+
"grad_norm": 1.4556822776794434,
|
| 682 |
+
"learning_rate": 4.490205523442518e-05,
|
| 683 |
+
"loss": 0.0516,
|
| 684 |
+
"step": 7200
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 6.0,
|
| 688 |
+
"eval_administration_accuracy": 0.9356889722743381,
|
| 689 |
+
"eval_administration_f1": 0.9317820245044639,
|
| 690 |
+
"eval_corruption_accuracy": 0.9579945799457995,
|
| 691 |
+
"eval_corruption_f1": 0.9568343721711066,
|
| 692 |
+
"eval_democracy_accuracy": 0.9589326662497394,
|
| 693 |
+
"eval_democracy_f1": 0.95607216308818,
|
| 694 |
+
"eval_development_accuracy": 0.9168230143839899,
|
| 695 |
+
"eval_development_f1": 0.9143401221807914,
|
| 696 |
+
"eval_economy_accuracy": 0.9308943089430894,
|
| 697 |
+
"eval_economy_f1": 0.9294738147048666,
|
| 698 |
+
"eval_education_accuracy": 0.9669585157390036,
|
| 699 |
+
"eval_education_f1": 0.965464450546169,
|
| 700 |
+
"eval_environment_accuracy": 0.9786324786324786,
|
| 701 |
+
"eval_environment_f1": 0.977778980162068,
|
| 702 |
+
"eval_instability_accuracy": 0.9441317490097978,
|
| 703 |
+
"eval_instability_f1": 0.9395705042517813,
|
| 704 |
+
"eval_leadership_accuracy": 0.8412549510110485,
|
| 705 |
+
"eval_leadership_f1": 0.8395230039371011,
|
| 706 |
+
"eval_loss": 0.26496145129203796,
|
| 707 |
+
"eval_overall_accuracy": 0.938372941421722,
|
| 708 |
+
"eval_overall_f1": 0.9362734744251768,
|
| 709 |
+
"eval_race_accuracy": 0.9541380029184907,
|
| 710 |
+
"eval_race_f1": 0.9519598766506976,
|
| 711 |
+
"eval_religion_accuracy": 0.9579945799457995,
|
| 712 |
+
"eval_religion_f1": 0.9568944771373694,
|
| 713 |
+
"eval_runtime": 10.8844,
|
| 714 |
+
"eval_safety_accuracy": 0.9170314780070877,
|
| 715 |
+
"eval_safety_f1": 0.915587903767528,
|
| 716 |
+
"eval_samples_per_second": 881.445,
|
| 717 |
+
"eval_steps_per_second": 55.125,
|
| 718 |
+
"step": 7200
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 6.083368070029179,
|
| 722 |
+
"grad_norm": 0.7286815643310547,
|
| 723 |
+
"learning_rate": 4.452740312566902e-05,
|
| 724 |
+
"loss": 0.0366,
|
| 725 |
+
"step": 7300
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 6.166736140058358,
|
| 729 |
+
"grad_norm": 2.7514851093292236,
|
| 730 |
+
"learning_rate": 4.4152751016912864e-05,
|
| 731 |
+
"loss": 0.0392,
|
| 732 |
+
"step": 7400
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 6.250104210087536,
|
| 736 |
+
"grad_norm": 1.0808284282684326,
|
| 737 |
+
"learning_rate": 4.377809890815671e-05,
|
| 738 |
+
"loss": 0.0365,
|
| 739 |
+
"step": 7500
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 6.333472280116715,
|
| 743 |
+
"grad_norm": 0.5677124261856079,
|
| 744 |
+
"learning_rate": 4.3403446799400555e-05,
|
| 745 |
+
"loss": 0.0439,
|
| 746 |
+
"step": 7600
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 6.416840350145894,
|
| 750 |
+
"grad_norm": 0.7869217991828918,
|
| 751 |
+
"learning_rate": 4.30287946906444e-05,
|
| 752 |
+
"loss": 0.034,
|
| 753 |
+
"step": 7700
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 6.500208420175073,
|
| 757 |
+
"grad_norm": 0.6197104454040527,
|
| 758 |
+
"learning_rate": 4.265414258188825e-05,
|
| 759 |
+
"loss": 0.0375,
|
| 760 |
+
"step": 7800
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 6.583576490204251,
|
| 764 |
+
"grad_norm": 1.126968264579773,
|
| 765 |
+
"learning_rate": 4.228323699421965e-05,
|
| 766 |
+
"loss": 0.0437,
|
| 767 |
+
"step": 7900
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 6.66694456023343,
|
| 771 |
+
"grad_norm": 0.6215279698371887,
|
| 772 |
+
"learning_rate": 4.190858488546349e-05,
|
| 773 |
+
"loss": 0.0413,
|
| 774 |
+
"step": 8000
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 6.750312630262609,
|
| 778 |
+
"grad_norm": 1.0289812088012695,
|
| 779 |
+
"learning_rate": 4.153393277670734e-05,
|
| 780 |
+
"loss": 0.0383,
|
| 781 |
+
"step": 8100
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 6.833680700291788,
|
| 785 |
+
"grad_norm": 1.2618753910064697,
|
| 786 |
+
"learning_rate": 4.1159280667951184e-05,
|
| 787 |
+
"loss": 0.037,
|
| 788 |
+
"step": 8200
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 6.917048770320967,
|
| 792 |
+
"grad_norm": 0.48271915316581726,
|
| 793 |
+
"learning_rate": 4.078462855919502e-05,
|
| 794 |
+
"loss": 0.0403,
|
| 795 |
+
"step": 8300
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 7.0,
|
| 799 |
+
"grad_norm": 1.4230297803878784,
|
| 800 |
+
"learning_rate": 4.040997645043887e-05,
|
| 801 |
+
"loss": 0.0421,
|
| 802 |
+
"step": 8400
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 7.0,
|
| 806 |
+
"eval_administration_accuracy": 0.9262038774233896,
|
| 807 |
+
"eval_administration_f1": 0.9271433160043757,
|
| 808 |
+
"eval_corruption_accuracy": 0.9570564936418595,
|
| 809 |
+
"eval_corruption_f1": 0.9567097147624816,
|
| 810 |
+
"eval_democracy_accuracy": 0.9531999166145507,
|
| 811 |
+
"eval_democracy_f1": 0.9521445432226068,
|
| 812 |
+
"eval_development_accuracy": 0.9057744423598082,
|
| 813 |
+
"eval_development_f1": 0.9056708555901262,
|
| 814 |
+
"eval_economy_accuracy": 0.9282885136543673,
|
| 815 |
+
"eval_economy_f1": 0.926345436006994,
|
| 816 |
+
"eval_education_accuracy": 0.962059620596206,
|
| 817 |
+
"eval_education_f1": 0.961835890625152,
|
| 818 |
+
"eval_environment_accuracy": 0.9805086512403586,
|
| 819 |
+
"eval_environment_f1": 0.9790877092918502,
|
| 820 |
+
"eval_instability_accuracy": 0.9382947675630603,
|
| 821 |
+
"eval_instability_f1": 0.9376909954205174,
|
| 822 |
+
"eval_leadership_accuracy": 0.8362518240567021,
|
| 823 |
+
"eval_leadership_f1": 0.8378707787677608,
|
| 824 |
+
"eval_loss": 0.2829027473926544,
|
| 825 |
+
"eval_overall_accuracy": 0.9348638037662428,
|
| 826 |
+
"eval_overall_f1": 0.9341947646002294,
|
| 827 |
+
"eval_race_accuracy": 0.9518449030644153,
|
| 828 |
+
"eval_race_f1": 0.9506769394120873,
|
| 829 |
+
"eval_religion_accuracy": 0.9585157390035439,
|
| 830 |
+
"eval_religion_f1": 0.9579338409216822,
|
| 831 |
+
"eval_runtime": 10.9473,
|
| 832 |
+
"eval_safety_accuracy": 0.9203668959766521,
|
| 833 |
+
"eval_safety_f1": 0.917227155177119,
|
| 834 |
+
"eval_samples_per_second": 876.38,
|
| 835 |
+
"eval_steps_per_second": 54.808,
|
| 836 |
+
"step": 8400
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 7.083368070029179,
|
| 840 |
+
"grad_norm": 0.5448200106620789,
|
| 841 |
+
"learning_rate": 4.0035324341682715e-05,
|
| 842 |
+
"loss": 0.0318,
|
| 843 |
+
"step": 8500
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 7.166736140058358,
|
| 847 |
+
"grad_norm": 0.2913892865180969,
|
| 848 |
+
"learning_rate": 3.966067223292656e-05,
|
| 849 |
+
"loss": 0.0286,
|
| 850 |
+
"step": 8600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 7.250104210087536,
|
| 854 |
+
"grad_norm": 0.5748587250709534,
|
| 855 |
+
"learning_rate": 3.9286020124170406e-05,
|
| 856 |
+
"loss": 0.0263,
|
| 857 |
+
"step": 8700
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 7.333472280116715,
|
| 861 |
+
"grad_norm": 1.5411152839660645,
|
| 862 |
+
"learning_rate": 3.891136801541425e-05,
|
| 863 |
+
"loss": 0.0284,
|
| 864 |
+
"step": 8800
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.416840350145894,
|
| 868 |
+
"grad_norm": 0.7248807549476624,
|
| 869 |
+
"learning_rate": 3.85367159066581e-05,
|
| 870 |
+
"loss": 0.0294,
|
| 871 |
+
"step": 8900
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.500208420175073,
|
| 875 |
+
"grad_norm": 1.4579120874404907,
|
| 876 |
+
"learning_rate": 3.8162063797901943e-05,
|
| 877 |
+
"loss": 0.0311,
|
| 878 |
+
"step": 9000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.583576490204251,
|
| 882 |
+
"grad_norm": 0.7706780433654785,
|
| 883 |
+
"learning_rate": 3.778741168914579e-05,
|
| 884 |
+
"loss": 0.0335,
|
| 885 |
+
"step": 9100
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 7.66694456023343,
|
| 889 |
+
"grad_norm": 2.0390424728393555,
|
| 890 |
+
"learning_rate": 3.7412759580389635e-05,
|
| 891 |
+
"loss": 0.031,
|
| 892 |
+
"step": 9200
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 7.750312630262609,
|
| 896 |
+
"grad_norm": 2.804685592651367,
|
| 897 |
+
"learning_rate": 3.7038107471633474e-05,
|
| 898 |
+
"loss": 0.0322,
|
| 899 |
+
"step": 9300
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 7.833680700291788,
|
| 903 |
+
"grad_norm": 1.1729462146759033,
|
| 904 |
+
"learning_rate": 3.666345536287732e-05,
|
| 905 |
+
"loss": 0.0318,
|
| 906 |
+
"step": 9400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 7.917048770320967,
|
| 910 |
+
"grad_norm": 0.402351051568985,
|
| 911 |
+
"learning_rate": 3.6288803254121165e-05,
|
| 912 |
+
"loss": 0.0328,
|
| 913 |
+
"step": 9500
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 8.0,
|
| 917 |
+
"grad_norm": 0.9323753714561462,
|
| 918 |
+
"learning_rate": 3.591415114536501e-05,
|
| 919 |
+
"loss": 0.0315,
|
| 920 |
+
"step": 9600
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 8.0,
|
| 924 |
+
"eval_administration_accuracy": 0.9343339587242027,
|
| 925 |
+
"eval_administration_f1": 0.9306710379016078,
|
| 926 |
+
"eval_corruption_accuracy": 0.9579945799457995,
|
| 927 |
+
"eval_corruption_f1": 0.956449462862611,
|
| 928 |
+
"eval_democracy_accuracy": 0.9521575984990619,
|
| 929 |
+
"eval_democracy_f1": 0.9518824877840592,
|
| 930 |
+
"eval_development_accuracy": 0.9153637690223057,
|
| 931 |
+
"eval_development_f1": 0.9115308958113573,
|
| 932 |
+
"eval_economy_accuracy": 0.9289139045236606,
|
| 933 |
+
"eval_economy_f1": 0.927826113707109,
|
| 934 |
+
"eval_education_accuracy": 0.9651865749426725,
|
| 935 |
+
"eval_education_f1": 0.9640713442699961,
|
| 936 |
+
"eval_environment_accuracy": 0.979153637690223,
|
| 937 |
+
"eval_environment_f1": 0.9780608387582607,
|
| 938 |
+
"eval_instability_accuracy": 0.9423598082134668,
|
| 939 |
+
"eval_instability_f1": 0.9402276314269008,
|
| 940 |
+
"eval_leadership_accuracy": 0.8358348968105066,
|
| 941 |
+
"eval_leadership_f1": 0.8373107575051231,
|
| 942 |
+
"eval_loss": 0.3004125952720642,
|
| 943 |
+
"eval_overall_accuracy": 0.9358018900701827,
|
| 944 |
+
"eval_overall_f1": 0.9347143795562429,
|
| 945 |
+
"eval_race_accuracy": 0.9542422347300397,
|
| 946 |
+
"eval_race_f1": 0.9531584962150744,
|
| 947 |
+
"eval_religion_accuracy": 0.9559099437148217,
|
| 948 |
+
"eval_religion_f1": 0.9557494733641548,
|
| 949 |
+
"eval_runtime": 11.0167,
|
| 950 |
+
"eval_safety_accuracy": 0.9081717740254326,
|
| 951 |
+
"eval_safety_f1": 0.9096340150686618,
|
| 952 |
+
"eval_samples_per_second": 870.856,
|
| 953 |
+
"eval_steps_per_second": 54.463,
|
| 954 |
+
"step": 9600
|
| 955 |
+
}
|
| 956 |
+
],
|
| 957 |
+
"logging_steps": 100,
|
| 958 |
+
"max_steps": 19184,
|
| 959 |
+
"num_input_tokens_seen": 0,
|
| 960 |
+
"num_train_epochs": 16,
|
| 961 |
+
"save_steps": 500,
|
| 962 |
+
"stateful_callbacks": {
|
| 963 |
+
"EarlyStoppingCallback": {
|
| 964 |
+
"args": {
|
| 965 |
+
"early_stopping_patience": 2,
|
| 966 |
+
"early_stopping_threshold": 0.0
|
| 967 |
+
},
|
| 968 |
+
"attributes": {
|
| 969 |
+
"early_stopping_patience_counter": 2
|
| 970 |
+
}
|
| 971 |
+
},
|
| 972 |
+
"TrainerControl": {
|
| 973 |
+
"args": {
|
| 974 |
+
"should_epoch_stop": false,
|
| 975 |
+
"should_evaluate": false,
|
| 976 |
+
"should_log": false,
|
| 977 |
+
"should_save": true,
|
| 978 |
+
"should_training_stop": true
|
| 979 |
+
},
|
| 980 |
+
"attributes": {}
|
| 981 |
+
}
|
| 982 |
+
},
|
| 983 |
+
"total_flos": 8.080213808303309e+16,
|
| 984 |
+
"train_batch_size": 16,
|
| 985 |
+
"trial_name": null,
|
| 986 |
+
"trial_params": {
|
| 987 |
+
"gradient_accumulation_steps": 2,
|
| 988 |
+
"learning_rate": 7e-05,
|
| 989 |
+
"num_train_epochs": 16
|
| 990 |
+
}
|
| 991 |
+
}
|
run-2/checkpoint-9600/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a417688ba3a38960888d219af94bafd1d9551df9005996513f2ac72dfc76c145
|
| 3 |
+
size 5368
|
run-3/checkpoint-2399/config.json
ADDED
|
@@ -0,0 +1,125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"BertForSequenceClassification"
|
| 4 |
+
],
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"classifier_dropout": null,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0",
|
| 13 |
+
"1": "LABEL_1",
|
| 14 |
+
"2": "LABEL_2",
|
| 15 |
+
"3": "LABEL_3",
|
| 16 |
+
"4": "LABEL_4",
|
| 17 |
+
"5": "LABEL_5",
|
| 18 |
+
"6": "LABEL_6",
|
| 19 |
+
"7": "LABEL_7",
|
| 20 |
+
"8": "LABEL_8",
|
| 21 |
+
"9": "LABEL_9",
|
| 22 |
+
"10": "LABEL_10",
|
| 23 |
+
"11": "LABEL_11",
|
| 24 |
+
"12": "LABEL_12",
|
| 25 |
+
"13": "LABEL_13",
|
| 26 |
+
"14": "LABEL_14",
|
| 27 |
+
"15": "LABEL_15",
|
| 28 |
+
"16": "LABEL_16",
|
| 29 |
+
"17": "LABEL_17",
|
| 30 |
+
"18": "LABEL_18",
|
| 31 |
+
"19": "LABEL_19",
|
| 32 |
+
"20": "LABEL_20",
|
| 33 |
+
"21": "LABEL_21",
|
| 34 |
+
"22": "LABEL_22",
|
| 35 |
+
"23": "LABEL_23",
|
| 36 |
+
"24": "LABEL_24",
|
| 37 |
+
"25": "LABEL_25",
|
| 38 |
+
"26": "LABEL_26",
|
| 39 |
+
"27": "LABEL_27",
|
| 40 |
+
"28": "LABEL_28",
|
| 41 |
+
"29": "LABEL_29",
|
| 42 |
+
"30": "LABEL_30",
|
| 43 |
+
"31": "LABEL_31",
|
| 44 |
+
"32": "LABEL_32",
|
| 45 |
+
"33": "LABEL_33",
|
| 46 |
+
"34": "LABEL_34",
|
| 47 |
+
"35": "LABEL_35",
|
| 48 |
+
"36": "LABEL_36",
|
| 49 |
+
"37": "LABEL_37",
|
| 50 |
+
"38": "LABEL_38",
|
| 51 |
+
"39": "LABEL_39",
|
| 52 |
+
"40": "LABEL_40",
|
| 53 |
+
"41": "LABEL_41",
|
| 54 |
+
"42": "LABEL_42",
|
| 55 |
+
"43": "LABEL_43",
|
| 56 |
+
"44": "LABEL_44",
|
| 57 |
+
"45": "LABEL_45",
|
| 58 |
+
"46": "LABEL_46",
|
| 59 |
+
"47": "LABEL_47"
|
| 60 |
+
},
|
| 61 |
+
"initializer_range": 0.02,
|
| 62 |
+
"intermediate_size": 3072,
|
| 63 |
+
"label2id": {
|
| 64 |
+
"LABEL_0": 0,
|
| 65 |
+
"LABEL_1": 1,
|
| 66 |
+
"LABEL_10": 10,
|
| 67 |
+
"LABEL_11": 11,
|
| 68 |
+
"LABEL_12": 12,
|
| 69 |
+
"LABEL_13": 13,
|
| 70 |
+
"LABEL_14": 14,
|
| 71 |
+
"LABEL_15": 15,
|
| 72 |
+
"LABEL_16": 16,
|
| 73 |
+
"LABEL_17": 17,
|
| 74 |
+
"LABEL_18": 18,
|
| 75 |
+
"LABEL_19": 19,
|
| 76 |
+
"LABEL_2": 2,
|
| 77 |
+
"LABEL_20": 20,
|
| 78 |
+
"LABEL_21": 21,
|
| 79 |
+
"LABEL_22": 22,
|
| 80 |
+
"LABEL_23": 23,
|
| 81 |
+
"LABEL_24": 24,
|
| 82 |
+
"LABEL_25": 25,
|
| 83 |
+
"LABEL_26": 26,
|
| 84 |
+
"LABEL_27": 27,
|
| 85 |
+
"LABEL_28": 28,
|
| 86 |
+
"LABEL_29": 29,
|
| 87 |
+
"LABEL_3": 3,
|
| 88 |
+
"LABEL_30": 30,
|
| 89 |
+
"LABEL_31": 31,
|
| 90 |
+
"LABEL_32": 32,
|
| 91 |
+
"LABEL_33": 33,
|
| 92 |
+
"LABEL_34": 34,
|
| 93 |
+
"LABEL_35": 35,
|
| 94 |
+
"LABEL_36": 36,
|
| 95 |
+
"LABEL_37": 37,
|
| 96 |
+
"LABEL_38": 38,
|
| 97 |
+
"LABEL_39": 39,
|
| 98 |
+
"LABEL_4": 4,
|
| 99 |
+
"LABEL_40": 40,
|
| 100 |
+
"LABEL_41": 41,
|
| 101 |
+
"LABEL_42": 42,
|
| 102 |
+
"LABEL_43": 43,
|
| 103 |
+
"LABEL_44": 44,
|
| 104 |
+
"LABEL_45": 45,
|
| 105 |
+
"LABEL_46": 46,
|
| 106 |
+
"LABEL_47": 47,
|
| 107 |
+
"LABEL_5": 5,
|
| 108 |
+
"LABEL_6": 6,
|
| 109 |
+
"LABEL_7": 7,
|
| 110 |
+
"LABEL_8": 8,
|
| 111 |
+
"LABEL_9": 9
|
| 112 |
+
},
|
| 113 |
+
"layer_norm_eps": 1e-12,
|
| 114 |
+
"max_position_embeddings": 512,
|
| 115 |
+
"model_type": "bert",
|
| 116 |
+
"num_attention_heads": 12,
|
| 117 |
+
"num_hidden_layers": 12,
|
| 118 |
+
"pad_token_id": 0,
|
| 119 |
+
"position_embedding_type": "absolute",
|
| 120 |
+
"torch_dtype": "float32",
|
| 121 |
+
"transformers_version": "4.50.3",
|
| 122 |
+
"type_vocab_size": 2,
|
| 123 |
+
"use_cache": true,
|
| 124 |
+
"vocab_size": 30522
|
| 125 |
+
}
|
run-3/checkpoint-2399/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8110a1a418d11311bbe6626de555c200d3eaaa703654697275995e81a68954e3
|
| 3 |
+
size 438100144
|
run-3/checkpoint-2399/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b1beca8967cfa6ad4956b3d3d14442dbb80815621011735a65eef74c188bb7c
|
| 3 |
+
size 876321402
|
run-3/checkpoint-2399/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03f6521411fa3fb31926db64cfd06b37e9d7c0607f2eec0bf790c73df4910854
|
| 3 |
+
size 14244
|
run-3/checkpoint-2399/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:682405df65ef7d615cbe34704d2dda54adc24ce53ed603b2d26612ee1de8aa21
|
| 3 |
+
size 988
|
run-3/checkpoint-2399/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:17c054bc86d448e79f0bbaa2e17a04e4966aa89a69f0a1f25078d59dd4884943
|
| 3 |
+
size 1064
|
run-3/checkpoint-2399/trainer_state.json
ADDED
|
@@ -0,0 +1,242 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2399,
|
| 3 |
+
"best_metric": 0.9219705720705208,
|
| 4 |
+
"best_model_checkpoint": "./results/run-3/checkpoint-2399",
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2399,
|
| 8 |
+
"is_hyper_param_search": true,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.041684035014589414,
|
| 14 |
+
"grad_norm": 1.7545671463012695,
|
| 15 |
+
"learning_rate": 1.386e-05,
|
| 16 |
+
"loss": 1.2954,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.08336807002917883,
|
| 21 |
+
"grad_norm": 0.7225435376167297,
|
| 22 |
+
"learning_rate": 2.7859999999999998e-05,
|
| 23 |
+
"loss": 0.5134,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.12505210504376824,
|
| 28 |
+
"grad_norm": 0.9463444948196411,
|
| 29 |
+
"learning_rate": 4.1859999999999996e-05,
|
| 30 |
+
"loss": 0.4399,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.16673614005835766,
|
| 35 |
+
"grad_norm": 0.8181287050247192,
|
| 36 |
+
"learning_rate": 5.586e-05,
|
| 37 |
+
"loss": 0.3608,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.20842017507294705,
|
| 42 |
+
"grad_norm": 1.4074561595916748,
|
| 43 |
+
"learning_rate": 6.986e-05,
|
| 44 |
+
"loss": 0.3303,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.25010421008753647,
|
| 49 |
+
"grad_norm": 1.232994556427002,
|
| 50 |
+
"learning_rate": 6.98170731707317e-05,
|
| 51 |
+
"loss": 0.3118,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.29178824510212586,
|
| 56 |
+
"grad_norm": 1.1151797771453857,
|
| 57 |
+
"learning_rate": 6.96341463414634e-05,
|
| 58 |
+
"loss": 0.3067,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.3334722801167153,
|
| 63 |
+
"grad_norm": 1.3453563451766968,
|
| 64 |
+
"learning_rate": 6.944937176644493e-05,
|
| 65 |
+
"loss": 0.2959,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3751563151313047,
|
| 70 |
+
"grad_norm": 1.3014110326766968,
|
| 71 |
+
"learning_rate": 6.926459719142646e-05,
|
| 72 |
+
"loss": 0.2803,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.4168403501458941,
|
| 77 |
+
"grad_norm": 2.4783546924591064,
|
| 78 |
+
"learning_rate": 6.907982261640798e-05,
|
| 79 |
+
"loss": 0.275,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.45852438516048355,
|
| 84 |
+
"grad_norm": 1.0991694927215576,
|
| 85 |
+
"learning_rate": 6.88950480413895e-05,
|
| 86 |
+
"loss": 0.2725,
|
| 87 |
+
"step": 1100
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.5002084201750729,
|
| 91 |
+
"grad_norm": 1.8534690141677856,
|
| 92 |
+
"learning_rate": 6.871027346637102e-05,
|
| 93 |
+
"loss": 0.2965,
|
| 94 |
+
"step": 1200
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.5418924551896623,
|
| 98 |
+
"grad_norm": 1.533751130104065,
|
| 99 |
+
"learning_rate": 6.852549889135254e-05,
|
| 100 |
+
"loss": 0.257,
|
| 101 |
+
"step": 1300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.5835764902042517,
|
| 105 |
+
"grad_norm": 1.39555025100708,
|
| 106 |
+
"learning_rate": 6.834072431633407e-05,
|
| 107 |
+
"loss": 0.2666,
|
| 108 |
+
"step": 1400
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.6252605252188412,
|
| 112 |
+
"grad_norm": 1.1627668142318726,
|
| 113 |
+
"learning_rate": 6.81559497413156e-05,
|
| 114 |
+
"loss": 0.2461,
|
| 115 |
+
"step": 1500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.6669445602334306,
|
| 119 |
+
"grad_norm": 1.319166660308838,
|
| 120 |
+
"learning_rate": 6.79711751662971e-05,
|
| 121 |
+
"loss": 0.242,
|
| 122 |
+
"step": 1600
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.70862859524802,
|
| 126 |
+
"grad_norm": 1.3535823822021484,
|
| 127 |
+
"learning_rate": 6.778640059127863e-05,
|
| 128 |
+
"loss": 0.2463,
|
| 129 |
+
"step": 1700
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.7503126302626094,
|
| 133 |
+
"grad_norm": 1.8609535694122314,
|
| 134 |
+
"learning_rate": 6.760162601626016e-05,
|
| 135 |
+
"loss": 0.2367,
|
| 136 |
+
"step": 1800
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.7919966652771988,
|
| 140 |
+
"grad_norm": 1.86128830909729,
|
| 141 |
+
"learning_rate": 6.741685144124168e-05,
|
| 142 |
+
"loss": 0.2449,
|
| 143 |
+
"step": 1900
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.8336807002917882,
|
| 147 |
+
"grad_norm": 1.8068938255310059,
|
| 148 |
+
"learning_rate": 6.723207686622321e-05,
|
| 149 |
+
"loss": 0.2329,
|
| 150 |
+
"step": 2000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.8753647353063777,
|
| 154 |
+
"grad_norm": 1.4341034889221191,
|
| 155 |
+
"learning_rate": 6.704730229120472e-05,
|
| 156 |
+
"loss": 0.2342,
|
| 157 |
+
"step": 2100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.9170487703209671,
|
| 161 |
+
"grad_norm": 0.7307045459747314,
|
| 162 |
+
"learning_rate": 6.686252771618624e-05,
|
| 163 |
+
"loss": 0.2358,
|
| 164 |
+
"step": 2200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.9587328053355565,
|
| 168 |
+
"grad_norm": 2.0308947563171387,
|
| 169 |
+
"learning_rate": 6.667775314116777e-05,
|
| 170 |
+
"loss": 0.2244,
|
| 171 |
+
"step": 2300
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.0,
|
| 175 |
+
"eval_administration_accuracy": 0.9317281634354805,
|
| 176 |
+
"eval_administration_f1": 0.9110301299328539,
|
| 177 |
+
"eval_corruption_accuracy": 0.9541380029184907,
|
| 178 |
+
"eval_corruption_f1": 0.9489574147600867,
|
| 179 |
+
"eval_democracy_accuracy": 0.9473629351678132,
|
| 180 |
+
"eval_democracy_f1": 0.935646230598331,
|
| 181 |
+
"eval_development_accuracy": 0.910464873879508,
|
| 182 |
+
"eval_development_f1": 0.9008646598644421,
|
| 183 |
+
"eval_economy_accuracy": 0.9235980821346675,
|
| 184 |
+
"eval_economy_f1": 0.9156600494130359,
|
| 185 |
+
"eval_education_accuracy": 0.9622680842193038,
|
| 186 |
+
"eval_education_f1": 0.9586233907920864,
|
| 187 |
+
"eval_environment_accuracy": 0.978319783197832,
|
| 188 |
+
"eval_environment_f1": 0.9770754665605412,
|
| 189 |
+
"eval_instability_accuracy": 0.9398582447362935,
|
| 190 |
+
"eval_instability_f1": 0.9299109706262821,
|
| 191 |
+
"eval_leadership_accuracy": 0.8081092349385033,
|
| 192 |
+
"eval_leadership_f1": 0.7818188134942813,
|
| 193 |
+
"eval_loss": 0.2264394313097,
|
| 194 |
+
"eval_overall_accuracy": 0.9310245987075256,
|
| 195 |
+
"eval_overall_f1": 0.9219705720705208,
|
| 196 |
+
"eval_race_accuracy": 0.9487179487179487,
|
| 197 |
+
"eval_race_f1": 0.9448562040025291,
|
| 198 |
+
"eval_religion_accuracy": 0.9527829893683553,
|
| 199 |
+
"eval_religion_f1": 0.950851997767183,
|
| 200 |
+
"eval_runtime": 11.1066,
|
| 201 |
+
"eval_safety_accuracy": 0.9149468417761101,
|
| 202 |
+
"eval_safety_f1": 0.9083515370345979,
|
| 203 |
+
"eval_samples_per_second": 863.809,
|
| 204 |
+
"eval_steps_per_second": 54.022,
|
| 205 |
+
"step": 2399
|
| 206 |
+
}
|
| 207 |
+
],
|
| 208 |
+
"logging_steps": 100,
|
| 209 |
+
"max_steps": 38384,
|
| 210 |
+
"num_input_tokens_seen": 0,
|
| 211 |
+
"num_train_epochs": 16,
|
| 212 |
+
"save_steps": 500,
|
| 213 |
+
"stateful_callbacks": {
|
| 214 |
+
"EarlyStoppingCallback": {
|
| 215 |
+
"args": {
|
| 216 |
+
"early_stopping_patience": 2,
|
| 217 |
+
"early_stopping_threshold": 0.0
|
| 218 |
+
},
|
| 219 |
+
"attributes": {
|
| 220 |
+
"early_stopping_patience_counter": 0
|
| 221 |
+
}
|
| 222 |
+
},
|
| 223 |
+
"TrainerControl": {
|
| 224 |
+
"args": {
|
| 225 |
+
"should_epoch_stop": false,
|
| 226 |
+
"should_evaluate": false,
|
| 227 |
+
"should_log": false,
|
| 228 |
+
"should_save": true,
|
| 229 |
+
"should_training_stop": false
|
| 230 |
+
},
|
| 231 |
+
"attributes": {}
|
| 232 |
+
}
|
| 233 |
+
},
|
| 234 |
+
"total_flos": 9686485853798400.0,
|
| 235 |
+
"train_batch_size": 16,
|
| 236 |
+
"trial_name": null,
|
| 237 |
+
"trial_params": {
|
| 238 |
+
"gradient_accumulation_steps": 1,
|
| 239 |
+
"learning_rate": 7e-05,
|
| 240 |
+
"num_train_epochs": 16
|
| 241 |
+
}
|
| 242 |
+
}
|
run-3/checkpoint-2399/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2574f00335e03e693d14802bd78255c32eb346c265bc7cfbd1b32046ac22108
|
| 3 |
+
size 5368
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2574f00335e03e693d14802bd78255c32eb346c265bc7cfbd1b32046ac22108
|
| 3 |
size 5368
|