Upload checkpoint-6505
Browse files- checkpoint-6505/config.json +60 -0
- checkpoint-6505/generation_config.json +243 -0
- checkpoint-6505/model.safetensors +3 -0
- checkpoint-6505/optimizer.pt +3 -0
- checkpoint-6505/preprocessor_config.json +14 -0
- checkpoint-6505/rng_state_0.pth +3 -0
- checkpoint-6505/rng_state_1.pth +3 -0
- checkpoint-6505/rng_state_2.pth +3 -0
- checkpoint-6505/rng_state_3.pth +3 -0
- checkpoint-6505/rng_state_4.pth +3 -0
- checkpoint-6505/rng_state_5.pth +3 -0
- checkpoint-6505/rng_state_6.pth +3 -0
- checkpoint-6505/rng_state_7.pth +3 -0
- checkpoint-6505/scheduler.pt +3 -0
- checkpoint-6505/trainer_state.json +944 -0
- checkpoint-6505/training_args.bin +3 -0
checkpoint-6505/config.json
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"activation_function": "gelu",
|
| 4 |
+
"apply_spec_augment": false,
|
| 5 |
+
"architectures": [
|
| 6 |
+
"WhisperForConditionalGeneration"
|
| 7 |
+
],
|
| 8 |
+
"attention_dropout": 0.0,
|
| 9 |
+
"begin_suppress_tokens": null,
|
| 10 |
+
"bos_token_id": 50257,
|
| 11 |
+
"classifier_proj_size": 256,
|
| 12 |
+
"d_model": 1024,
|
| 13 |
+
"decoder_attention_heads": 16,
|
| 14 |
+
"decoder_ffn_dim": 4096,
|
| 15 |
+
"decoder_layerdrop": 0.0,
|
| 16 |
+
"decoder_layers": 24,
|
| 17 |
+
"decoder_start_token_id": 50258,
|
| 18 |
+
"dropout": 0.0,
|
| 19 |
+
"dtype": "float32",
|
| 20 |
+
"encoder_attention_heads": 16,
|
| 21 |
+
"encoder_ffn_dim": 4096,
|
| 22 |
+
"encoder_layerdrop": 0.0,
|
| 23 |
+
"encoder_layers": 24,
|
| 24 |
+
"eos_token_id": 50257,
|
| 25 |
+
"forced_decoder_ids": [
|
| 26 |
+
[
|
| 27 |
+
1,
|
| 28 |
+
50259
|
| 29 |
+
],
|
| 30 |
+
[
|
| 31 |
+
2,
|
| 32 |
+
50359
|
| 33 |
+
],
|
| 34 |
+
[
|
| 35 |
+
3,
|
| 36 |
+
50363
|
| 37 |
+
]
|
| 38 |
+
],
|
| 39 |
+
"init_std": 0.02,
|
| 40 |
+
"is_encoder_decoder": true,
|
| 41 |
+
"mask_feature_length": 10,
|
| 42 |
+
"mask_feature_min_masks": 0,
|
| 43 |
+
"mask_feature_prob": 0.0,
|
| 44 |
+
"mask_time_length": 10,
|
| 45 |
+
"mask_time_min_masks": 2,
|
| 46 |
+
"mask_time_prob": 0.05,
|
| 47 |
+
"max_source_positions": 1500,
|
| 48 |
+
"max_target_positions": 448,
|
| 49 |
+
"median_filter_width": 7,
|
| 50 |
+
"model_type": "whisper",
|
| 51 |
+
"num_mel_bins": 80,
|
| 52 |
+
"pad_token_id": 50257,
|
| 53 |
+
"scale_embedding": false,
|
| 54 |
+
"suppress_tokens": null,
|
| 55 |
+
"tie_word_embeddings": true,
|
| 56 |
+
"transformers_version": "5.4.0",
|
| 57 |
+
"use_cache": false,
|
| 58 |
+
"use_weighted_layer_sum": false,
|
| 59 |
+
"vocab_size": 51865
|
| 60 |
+
}
|
checkpoint-6505/generation_config.json
ADDED
|
@@ -0,0 +1,243 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alignment_heads": [
|
| 3 |
+
[
|
| 4 |
+
13,
|
| 5 |
+
15
|
| 6 |
+
],
|
| 7 |
+
[
|
| 8 |
+
15,
|
| 9 |
+
4
|
| 10 |
+
],
|
| 11 |
+
[
|
| 12 |
+
15,
|
| 13 |
+
15
|
| 14 |
+
],
|
| 15 |
+
[
|
| 16 |
+
16,
|
| 17 |
+
1
|
| 18 |
+
],
|
| 19 |
+
[
|
| 20 |
+
20,
|
| 21 |
+
0
|
| 22 |
+
],
|
| 23 |
+
[
|
| 24 |
+
23,
|
| 25 |
+
4
|
| 26 |
+
]
|
| 27 |
+
],
|
| 28 |
+
"begin_suppress_tokens": [
|
| 29 |
+
220,
|
| 30 |
+
50257
|
| 31 |
+
],
|
| 32 |
+
"bos_token_id": 50257,
|
| 33 |
+
"decoder_start_token_id": 50258,
|
| 34 |
+
"eos_token_id": [
|
| 35 |
+
50257
|
| 36 |
+
],
|
| 37 |
+
"forced_decoder_ids": null,
|
| 38 |
+
"is_multilingual": true,
|
| 39 |
+
"lang_to_id": {
|
| 40 |
+
"<|af|>": 50327,
|
| 41 |
+
"<|am|>": 50334,
|
| 42 |
+
"<|ar|>": 50272,
|
| 43 |
+
"<|as|>": 50350,
|
| 44 |
+
"<|az|>": 50304,
|
| 45 |
+
"<|ba|>": 50355,
|
| 46 |
+
"<|be|>": 50330,
|
| 47 |
+
"<|bg|>": 50292,
|
| 48 |
+
"<|bn|>": 50302,
|
| 49 |
+
"<|bo|>": 50347,
|
| 50 |
+
"<|br|>": 50309,
|
| 51 |
+
"<|bs|>": 50315,
|
| 52 |
+
"<|ca|>": 50270,
|
| 53 |
+
"<|cs|>": 50283,
|
| 54 |
+
"<|cy|>": 50297,
|
| 55 |
+
"<|da|>": 50285,
|
| 56 |
+
"<|de|>": 50261,
|
| 57 |
+
"<|el|>": 50281,
|
| 58 |
+
"<|en|>": 50259,
|
| 59 |
+
"<|es|>": 50262,
|
| 60 |
+
"<|et|>": 50307,
|
| 61 |
+
"<|eu|>": 50310,
|
| 62 |
+
"<|fa|>": 50300,
|
| 63 |
+
"<|fi|>": 50277,
|
| 64 |
+
"<|fo|>": 50338,
|
| 65 |
+
"<|fr|>": 50265,
|
| 66 |
+
"<|gl|>": 50319,
|
| 67 |
+
"<|gu|>": 50333,
|
| 68 |
+
"<|haw|>": 50352,
|
| 69 |
+
"<|ha|>": 50354,
|
| 70 |
+
"<|he|>": 50279,
|
| 71 |
+
"<|hi|>": 50276,
|
| 72 |
+
"<|hr|>": 50291,
|
| 73 |
+
"<|ht|>": 50339,
|
| 74 |
+
"<|hu|>": 50286,
|
| 75 |
+
"<|hy|>": 50312,
|
| 76 |
+
"<|id|>": 50275,
|
| 77 |
+
"<|is|>": 50311,
|
| 78 |
+
"<|it|>": 50274,
|
| 79 |
+
"<|ja|>": 50266,
|
| 80 |
+
"<|jw|>": 50356,
|
| 81 |
+
"<|ka|>": 50329,
|
| 82 |
+
"<|kk|>": 50316,
|
| 83 |
+
"<|km|>": 50323,
|
| 84 |
+
"<|kn|>": 50306,
|
| 85 |
+
"<|ko|>": 50264,
|
| 86 |
+
"<|la|>": 50294,
|
| 87 |
+
"<|lb|>": 50345,
|
| 88 |
+
"<|ln|>": 50353,
|
| 89 |
+
"<|lo|>": 50336,
|
| 90 |
+
"<|lt|>": 50293,
|
| 91 |
+
"<|lv|>": 50301,
|
| 92 |
+
"<|mg|>": 50349,
|
| 93 |
+
"<|mi|>": 50295,
|
| 94 |
+
"<|mk|>": 50308,
|
| 95 |
+
"<|ml|>": 50296,
|
| 96 |
+
"<|mn|>": 50314,
|
| 97 |
+
"<|mr|>": 50320,
|
| 98 |
+
"<|ms|>": 50282,
|
| 99 |
+
"<|mt|>": 50343,
|
| 100 |
+
"<|my|>": 50346,
|
| 101 |
+
"<|ne|>": 50313,
|
| 102 |
+
"<|nl|>": 50271,
|
| 103 |
+
"<|nn|>": 50342,
|
| 104 |
+
"<|no|>": 50288,
|
| 105 |
+
"<|oc|>": 50328,
|
| 106 |
+
"<|pa|>": 50321,
|
| 107 |
+
"<|pl|>": 50269,
|
| 108 |
+
"<|ps|>": 50340,
|
| 109 |
+
"<|pt|>": 50267,
|
| 110 |
+
"<|ro|>": 50284,
|
| 111 |
+
"<|ru|>": 50263,
|
| 112 |
+
"<|sa|>": 50344,
|
| 113 |
+
"<|sd|>": 50332,
|
| 114 |
+
"<|si|>": 50322,
|
| 115 |
+
"<|sk|>": 50298,
|
| 116 |
+
"<|sl|>": 50305,
|
| 117 |
+
"<|sn|>": 50324,
|
| 118 |
+
"<|so|>": 50326,
|
| 119 |
+
"<|sq|>": 50317,
|
| 120 |
+
"<|sr|>": 50303,
|
| 121 |
+
"<|su|>": 50357,
|
| 122 |
+
"<|sv|>": 50273,
|
| 123 |
+
"<|sw|>": 50318,
|
| 124 |
+
"<|ta|>": 50287,
|
| 125 |
+
"<|te|>": 50299,
|
| 126 |
+
"<|tg|>": 50331,
|
| 127 |
+
"<|th|>": 50289,
|
| 128 |
+
"<|tk|>": 50341,
|
| 129 |
+
"<|tl|>": 50348,
|
| 130 |
+
"<|tr|>": 50268,
|
| 131 |
+
"<|tt|>": 50351,
|
| 132 |
+
"<|uk|>": 50280,
|
| 133 |
+
"<|ur|>": 50290,
|
| 134 |
+
"<|uz|>": 50337,
|
| 135 |
+
"<|vi|>": 50278,
|
| 136 |
+
"<|yi|>": 50335,
|
| 137 |
+
"<|yo|>": 50325,
|
| 138 |
+
"<|zh|>": 50260
|
| 139 |
+
},
|
| 140 |
+
"language": "Hindi",
|
| 141 |
+
"max_initial_timestamp_index": 50,
|
| 142 |
+
"max_length": 448,
|
| 143 |
+
"no_timestamps_token_id": 50363,
|
| 144 |
+
"pad_token_id": 50257,
|
| 145 |
+
"prev_sot_token_id": 50361,
|
| 146 |
+
"return_timestamps": false,
|
| 147 |
+
"suppress_tokens": [
|
| 148 |
+
1,
|
| 149 |
+
2,
|
| 150 |
+
7,
|
| 151 |
+
8,
|
| 152 |
+
9,
|
| 153 |
+
10,
|
| 154 |
+
14,
|
| 155 |
+
25,
|
| 156 |
+
26,
|
| 157 |
+
27,
|
| 158 |
+
28,
|
| 159 |
+
29,
|
| 160 |
+
31,
|
| 161 |
+
58,
|
| 162 |
+
59,
|
| 163 |
+
60,
|
| 164 |
+
61,
|
| 165 |
+
62,
|
| 166 |
+
63,
|
| 167 |
+
90,
|
| 168 |
+
91,
|
| 169 |
+
92,
|
| 170 |
+
93,
|
| 171 |
+
359,
|
| 172 |
+
503,
|
| 173 |
+
522,
|
| 174 |
+
542,
|
| 175 |
+
873,
|
| 176 |
+
893,
|
| 177 |
+
902,
|
| 178 |
+
918,
|
| 179 |
+
922,
|
| 180 |
+
931,
|
| 181 |
+
1350,
|
| 182 |
+
1853,
|
| 183 |
+
1982,
|
| 184 |
+
2460,
|
| 185 |
+
2627,
|
| 186 |
+
3246,
|
| 187 |
+
3253,
|
| 188 |
+
3268,
|
| 189 |
+
3536,
|
| 190 |
+
3846,
|
| 191 |
+
3961,
|
| 192 |
+
4183,
|
| 193 |
+
4667,
|
| 194 |
+
6585,
|
| 195 |
+
6647,
|
| 196 |
+
7273,
|
| 197 |
+
9061,
|
| 198 |
+
9383,
|
| 199 |
+
10428,
|
| 200 |
+
10929,
|
| 201 |
+
11938,
|
| 202 |
+
12033,
|
| 203 |
+
12331,
|
| 204 |
+
12562,
|
| 205 |
+
13793,
|
| 206 |
+
14157,
|
| 207 |
+
14635,
|
| 208 |
+
15265,
|
| 209 |
+
15618,
|
| 210 |
+
16553,
|
| 211 |
+
16604,
|
| 212 |
+
18362,
|
| 213 |
+
18956,
|
| 214 |
+
20075,
|
| 215 |
+
21675,
|
| 216 |
+
22520,
|
| 217 |
+
26130,
|
| 218 |
+
26161,
|
| 219 |
+
26435,
|
| 220 |
+
28279,
|
| 221 |
+
29464,
|
| 222 |
+
31650,
|
| 223 |
+
32302,
|
| 224 |
+
32470,
|
| 225 |
+
36865,
|
| 226 |
+
42863,
|
| 227 |
+
47425,
|
| 228 |
+
49870,
|
| 229 |
+
50254,
|
| 230 |
+
50258,
|
| 231 |
+
50358,
|
| 232 |
+
50359,
|
| 233 |
+
50360,
|
| 234 |
+
50361,
|
| 235 |
+
50362
|
| 236 |
+
],
|
| 237 |
+
"task": "transcribe",
|
| 238 |
+
"task_to_id": {
|
| 239 |
+
"transcribe": 50359,
|
| 240 |
+
"translate": 50358
|
| 241 |
+
},
|
| 242 |
+
"transformers_version": "5.4.0"
|
| 243 |
+
}
|
checkpoint-6505/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72886def25a6ccfae0b50846544239f66fabf9ad2e5c28db9af6a6eed3e00bd6
|
| 3 |
+
size 3055544304
|
checkpoint-6505/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02eb900a05090a63d7dbad8c01871a600509410b101f9f74c85f703cbea92eba
|
| 3 |
+
size 6111697804
|
checkpoint-6505/preprocessor_config.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"chunk_length": 30,
|
| 3 |
+
"dither": 0.0,
|
| 4 |
+
"feature_extractor_type": "WhisperFeatureExtractor",
|
| 5 |
+
"feature_size": 80,
|
| 6 |
+
"hop_length": 160,
|
| 7 |
+
"n_fft": 400,
|
| 8 |
+
"n_samples": 480000,
|
| 9 |
+
"nb_max_frames": 3000,
|
| 10 |
+
"padding_side": "right",
|
| 11 |
+
"padding_value": 0.0,
|
| 12 |
+
"return_attention_mask": false,
|
| 13 |
+
"sampling_rate": 16000
|
| 14 |
+
}
|
checkpoint-6505/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:478a6eda5f4b9fe3812967c69e7f67171003588fe211f7b2b28a8b23201a6613
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a1ade1e5a198fb0201be5260aebce5d010a7d9a519b5149c4f68b571fed2ef36
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cc643bade276d6ced17cd04bf35f7f1675ddb96754e2ae67065957e9fb4c1232
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d1b275e90030a777ac095504d821baa0c73f32ffd8acf72c5c5933b1a5f274e3
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb4abbfc90417bb9873aeb9aab62645e73caa8708fea2615b1eca028cb7fb7ed
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c1aa2d73ee1f2342c29a47102733749842947f7cb6aa0956aefae54b2ac9562
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81c6ea778e3fba14052956593599f833e903241a60c1584fd82f0112267d0410
|
| 3 |
+
size 16389
|
checkpoint-6505/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec86fa9e6cf97888571b63ec02cbf2f0350c5f3b9df6fe66d9d98f57e6f58595
|
| 3 |
+
size 16389
|
checkpoint-6505/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63087e8eab2569faf42553423b3d1c129508fd7b5b910729704a8337190c15a5
|
| 3 |
+
size 1465
|
checkpoint-6505/trainer_state.json
ADDED
|
@@ -0,0 +1,944 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 6505,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.03843197540353574,
|
| 14 |
+
"grad_norm": 90.69124603271484,
|
| 15 |
+
"learning_rate": 1.4e-07,
|
| 16 |
+
"loss": 4.672296447753906,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.07686395080707148,
|
| 21 |
+
"grad_norm": 15.411819458007812,
|
| 22 |
+
"learning_rate": 2.8285714285714286e-07,
|
| 23 |
+
"loss": 3.5202178955078125,
|
| 24 |
+
"step": 100
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.11529592621060722,
|
| 28 |
+
"grad_norm": 9.78220272064209,
|
| 29 |
+
"learning_rate": 4.257142857142857e-07,
|
| 30 |
+
"loss": 2.5221026611328123,
|
| 31 |
+
"step": 150
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.15372790161414296,
|
| 35 |
+
"grad_norm": 9.672887802124023,
|
| 36 |
+
"learning_rate": 5.685714285714286e-07,
|
| 37 |
+
"loss": 2.1046574401855467,
|
| 38 |
+
"step": 200
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.1921598770176787,
|
| 42 |
+
"grad_norm": 8.549321174621582,
|
| 43 |
+
"learning_rate": 7.114285714285714e-07,
|
| 44 |
+
"loss": 1.9424900817871094,
|
| 45 |
+
"step": 250
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.23059185242121444,
|
| 49 |
+
"grad_norm": 9.276304244995117,
|
| 50 |
+
"learning_rate": 8.542857142857142e-07,
|
| 51 |
+
"loss": 1.7669134521484375,
|
| 52 |
+
"step": 300
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.2690238278247502,
|
| 56 |
+
"grad_norm": 9.018757820129395,
|
| 57 |
+
"learning_rate": 9.97142857142857e-07,
|
| 58 |
+
"loss": 1.621932830810547,
|
| 59 |
+
"step": 350
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.3074558032282859,
|
| 63 |
+
"grad_norm": 9.333130836486816,
|
| 64 |
+
"learning_rate": 1.1399999999999999e-06,
|
| 65 |
+
"loss": 1.407852783203125,
|
| 66 |
+
"step": 400
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.3458877786318217,
|
| 70 |
+
"grad_norm": 5.959807395935059,
|
| 71 |
+
"learning_rate": 1.282857142857143e-06,
|
| 72 |
+
"loss": 1.1452353668212891,
|
| 73 |
+
"step": 450
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.3843197540353574,
|
| 77 |
+
"grad_norm": 5.901744365692139,
|
| 78 |
+
"learning_rate": 1.4257142857142857e-06,
|
| 79 |
+
"loss": 0.9521345520019531,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.42275172943889316,
|
| 84 |
+
"grad_norm": 5.9748358726501465,
|
| 85 |
+
"learning_rate": 1.5685714285714283e-06,
|
| 86 |
+
"loss": 0.8781891632080078,
|
| 87 |
+
"step": 550
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.4611837048424289,
|
| 91 |
+
"grad_norm": 5.292792797088623,
|
| 92 |
+
"learning_rate": 1.7114285714285714e-06,
|
| 93 |
+
"loss": 0.9045460510253907,
|
| 94 |
+
"step": 600
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.49961568024596464,
|
| 98 |
+
"grad_norm": 5.277599334716797,
|
| 99 |
+
"learning_rate": 1.8542857142857142e-06,
|
| 100 |
+
"loss": 0.901295166015625,
|
| 101 |
+
"step": 650
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.5380476556495004,
|
| 105 |
+
"grad_norm": 4.598923683166504,
|
| 106 |
+
"learning_rate": 1.997142857142857e-06,
|
| 107 |
+
"loss": 0.9063746643066406,
|
| 108 |
+
"step": 700
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.5764796310530361,
|
| 112 |
+
"grad_norm": 5.642742156982422,
|
| 113 |
+
"learning_rate": 1.992038992688871e-06,
|
| 114 |
+
"loss": 0.85524658203125,
|
| 115 |
+
"step": 750
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.6149116064565718,
|
| 119 |
+
"grad_norm": 5.199263095855713,
|
| 120 |
+
"learning_rate": 1.98391551584078e-06,
|
| 121 |
+
"loss": 0.8806074523925781,
|
| 122 |
+
"step": 800
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.6533435818601077,
|
| 126 |
+
"grad_norm": 4.932873249053955,
|
| 127 |
+
"learning_rate": 1.9757920389926886e-06,
|
| 128 |
+
"loss": 0.8675950622558594,
|
| 129 |
+
"step": 850
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.6917755572636434,
|
| 133 |
+
"grad_norm": 5.1567463874816895,
|
| 134 |
+
"learning_rate": 1.9676685621445978e-06,
|
| 135 |
+
"loss": 0.8581599426269532,
|
| 136 |
+
"step": 900
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.7302075326671791,
|
| 140 |
+
"grad_norm": 5.477665901184082,
|
| 141 |
+
"learning_rate": 1.959545085296507e-06,
|
| 142 |
+
"loss": 0.8491453552246093,
|
| 143 |
+
"step": 950
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.7686395080707148,
|
| 147 |
+
"grad_norm": 4.780323505401611,
|
| 148 |
+
"learning_rate": 1.951421608448416e-06,
|
| 149 |
+
"loss": 0.85265380859375,
|
| 150 |
+
"step": 1000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.8070714834742506,
|
| 154 |
+
"grad_norm": 4.965339660644531,
|
| 155 |
+
"learning_rate": 1.9432981316003246e-06,
|
| 156 |
+
"loss": 0.8527495574951172,
|
| 157 |
+
"step": 1050
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.8455034588777863,
|
| 161 |
+
"grad_norm": 4.790028095245361,
|
| 162 |
+
"learning_rate": 1.9351746547522337e-06,
|
| 163 |
+
"loss": 0.8524763488769531,
|
| 164 |
+
"step": 1100
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.883935434281322,
|
| 168 |
+
"grad_norm": 5.199764251708984,
|
| 169 |
+
"learning_rate": 1.927051177904143e-06,
|
| 170 |
+
"loss": 0.8460150146484375,
|
| 171 |
+
"step": 1150
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.9223674096848578,
|
| 175 |
+
"grad_norm": 4.704173564910889,
|
| 176 |
+
"learning_rate": 1.918927701056052e-06,
|
| 177 |
+
"loss": 0.8334638977050781,
|
| 178 |
+
"step": 1200
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.9607993850883936,
|
| 182 |
+
"grad_norm": 4.4041523933410645,
|
| 183 |
+
"learning_rate": 1.910804224207961e-06,
|
| 184 |
+
"loss": 0.8634848022460937,
|
| 185 |
+
"step": 1250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.9992313604919293,
|
| 189 |
+
"grad_norm": 5.256194591522217,
|
| 190 |
+
"learning_rate": 1.9026807473598697e-06,
|
| 191 |
+
"loss": 0.8514776611328125,
|
| 192 |
+
"step": 1300
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.037663335895465,
|
| 196 |
+
"grad_norm": 6.466175079345703,
|
| 197 |
+
"learning_rate": 1.894557270511779e-06,
|
| 198 |
+
"loss": 0.7951304626464843,
|
| 199 |
+
"step": 1350
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.0760953112990008,
|
| 203 |
+
"grad_norm": 5.014453411102295,
|
| 204 |
+
"learning_rate": 1.8864337936636881e-06,
|
| 205 |
+
"loss": 0.7868093872070312,
|
| 206 |
+
"step": 1400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.1145272867025364,
|
| 210 |
+
"grad_norm": 4.826796054840088,
|
| 211 |
+
"learning_rate": 1.8783103168155968e-06,
|
| 212 |
+
"loss": 0.7673994445800781,
|
| 213 |
+
"step": 1450
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.1529592621060722,
|
| 217 |
+
"grad_norm": 5.103256702423096,
|
| 218 |
+
"learning_rate": 1.8701868399675061e-06,
|
| 219 |
+
"loss": 0.7927820587158203,
|
| 220 |
+
"step": 1500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 1.191391237509608,
|
| 224 |
+
"grad_norm": 5.426547050476074,
|
| 225 |
+
"learning_rate": 1.8620633631194148e-06,
|
| 226 |
+
"loss": 0.7597624206542969,
|
| 227 |
+
"step": 1550
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.2298232129131437,
|
| 231 |
+
"grad_norm": 5.586554050445557,
|
| 232 |
+
"learning_rate": 1.853939886271324e-06,
|
| 233 |
+
"loss": 0.7939593505859375,
|
| 234 |
+
"step": 1600
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 1.2682551883166795,
|
| 238 |
+
"grad_norm": 4.912991046905518,
|
| 239 |
+
"learning_rate": 1.8458164094232332e-06,
|
| 240 |
+
"loss": 0.7834192657470703,
|
| 241 |
+
"step": 1650
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.3066871637202153,
|
| 245 |
+
"grad_norm": 4.797924995422363,
|
| 246 |
+
"learning_rate": 1.837692932575142e-06,
|
| 247 |
+
"loss": 0.777972183227539,
|
| 248 |
+
"step": 1700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.345119139123751,
|
| 252 |
+
"grad_norm": 5.317209243774414,
|
| 253 |
+
"learning_rate": 1.829569455727051e-06,
|
| 254 |
+
"loss": 0.7747420501708985,
|
| 255 |
+
"step": 1750
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.3835511145272867,
|
| 259 |
+
"grad_norm": 4.722371578216553,
|
| 260 |
+
"learning_rate": 1.82144597887896e-06,
|
| 261 |
+
"loss": 0.7762690734863281,
|
| 262 |
+
"step": 1800
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.4219830899308223,
|
| 266 |
+
"grad_norm": 4.524418354034424,
|
| 267 |
+
"learning_rate": 1.813322502030869e-06,
|
| 268 |
+
"loss": 0.7724872589111328,
|
| 269 |
+
"step": 1850
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.4604150653343582,
|
| 273 |
+
"grad_norm": 4.875087738037109,
|
| 274 |
+
"learning_rate": 1.8051990251827781e-06,
|
| 275 |
+
"loss": 0.7988505554199219,
|
| 276 |
+
"step": 1900
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.498847040737894,
|
| 280 |
+
"grad_norm": 6.139016151428223,
|
| 281 |
+
"learning_rate": 1.797075548334687e-06,
|
| 282 |
+
"loss": 0.7693194580078125,
|
| 283 |
+
"step": 1950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.5372790161414298,
|
| 287 |
+
"grad_norm": 4.9258551597595215,
|
| 288 |
+
"learning_rate": 1.7889520714865961e-06,
|
| 289 |
+
"loss": 0.7665122985839844,
|
| 290 |
+
"step": 2000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.5757109915449654,
|
| 294 |
+
"grad_norm": 4.519026756286621,
|
| 295 |
+
"learning_rate": 1.780828594638505e-06,
|
| 296 |
+
"loss": 0.7608248901367187,
|
| 297 |
+
"step": 2050
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.614142966948501,
|
| 301 |
+
"grad_norm": 4.66514778137207,
|
| 302 |
+
"learning_rate": 1.7727051177904141e-06,
|
| 303 |
+
"loss": 0.7740484619140625,
|
| 304 |
+
"step": 2100
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.6525749423520368,
|
| 308 |
+
"grad_norm": 4.593281269073486,
|
| 309 |
+
"learning_rate": 1.7645816409423232e-06,
|
| 310 |
+
"loss": 0.7677935791015625,
|
| 311 |
+
"step": 2150
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.6910069177555727,
|
| 315 |
+
"grad_norm": 4.811053276062012,
|
| 316 |
+
"learning_rate": 1.756458164094232e-06,
|
| 317 |
+
"loss": 0.7720513916015626,
|
| 318 |
+
"step": 2200
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 1.7294388931591085,
|
| 322 |
+
"grad_norm": 4.175657749176025,
|
| 323 |
+
"learning_rate": 1.7483346872461412e-06,
|
| 324 |
+
"loss": 0.7719061279296875,
|
| 325 |
+
"step": 2250
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 1.767870868562644,
|
| 329 |
+
"grad_norm": 4.676957130432129,
|
| 330 |
+
"learning_rate": 1.74021121039805e-06,
|
| 331 |
+
"loss": 0.7846797180175781,
|
| 332 |
+
"step": 2300
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.80630284396618,
|
| 336 |
+
"grad_norm": 4.920013904571533,
|
| 337 |
+
"learning_rate": 1.7320877335499592e-06,
|
| 338 |
+
"loss": 0.7882473754882813,
|
| 339 |
+
"step": 2350
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.8447348193697155,
|
| 343 |
+
"grad_norm": 3.9824724197387695,
|
| 344 |
+
"learning_rate": 1.7239642567018683e-06,
|
| 345 |
+
"loss": 0.7606929016113281,
|
| 346 |
+
"step": 2400
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 1.8831667947732513,
|
| 350 |
+
"grad_norm": 5.4492692947387695,
|
| 351 |
+
"learning_rate": 1.7158407798537772e-06,
|
| 352 |
+
"loss": 0.7734042358398437,
|
| 353 |
+
"step": 2450
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.9215987701767872,
|
| 357 |
+
"grad_norm": 4.610161781311035,
|
| 358 |
+
"learning_rate": 1.7077173030056863e-06,
|
| 359 |
+
"loss": 0.7572845458984375,
|
| 360 |
+
"step": 2500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.960030745580323,
|
| 364 |
+
"grad_norm": 4.945580005645752,
|
| 365 |
+
"learning_rate": 1.6995938261575952e-06,
|
| 366 |
+
"loss": 0.7701009368896484,
|
| 367 |
+
"step": 2550
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.9984627209838586,
|
| 371 |
+
"grad_norm": 4.540768146514893,
|
| 372 |
+
"learning_rate": 1.6914703493095043e-06,
|
| 373 |
+
"loss": 0.7705160522460938,
|
| 374 |
+
"step": 2600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.036894696387394,
|
| 378 |
+
"grad_norm": 4.4660139083862305,
|
| 379 |
+
"learning_rate": 1.6833468724614134e-06,
|
| 380 |
+
"loss": 0.7190726470947265,
|
| 381 |
+
"step": 2650
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 2.07532667179093,
|
| 385 |
+
"grad_norm": 4.253429889678955,
|
| 386 |
+
"learning_rate": 1.6752233956133223e-06,
|
| 387 |
+
"loss": 0.6725291442871094,
|
| 388 |
+
"step": 2700
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 2.113758647194466,
|
| 392 |
+
"grad_norm": 4.271991729736328,
|
| 393 |
+
"learning_rate": 1.6670999187652314e-06,
|
| 394 |
+
"loss": 0.6934330749511719,
|
| 395 |
+
"step": 2750
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 2.1521906225980016,
|
| 399 |
+
"grad_norm": 5.311951160430908,
|
| 400 |
+
"learning_rate": 1.6589764419171403e-06,
|
| 401 |
+
"loss": 0.6960661315917969,
|
| 402 |
+
"step": 2800
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 2.1906225980015375,
|
| 406 |
+
"grad_norm": 4.163394927978516,
|
| 407 |
+
"learning_rate": 1.6508529650690494e-06,
|
| 408 |
+
"loss": 0.6919083404541015,
|
| 409 |
+
"step": 2850
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.229054573405073,
|
| 413 |
+
"grad_norm": 4.293502330780029,
|
| 414 |
+
"learning_rate": 1.6427294882209585e-06,
|
| 415 |
+
"loss": 0.7053424072265625,
|
| 416 |
+
"step": 2900
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.2674865488086087,
|
| 420 |
+
"grad_norm": 4.649949550628662,
|
| 421 |
+
"learning_rate": 1.6346060113728674e-06,
|
| 422 |
+
"loss": 0.6945845031738281,
|
| 423 |
+
"step": 2950
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.3059185242121445,
|
| 427 |
+
"grad_norm": 4.741189002990723,
|
| 428 |
+
"learning_rate": 1.6264825345247765e-06,
|
| 429 |
+
"loss": 0.6965169525146484,
|
| 430 |
+
"step": 3000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.3443504996156803,
|
| 434 |
+
"grad_norm": 4.3140482902526855,
|
| 435 |
+
"learning_rate": 1.6183590576766854e-06,
|
| 436 |
+
"loss": 0.7144229125976562,
|
| 437 |
+
"step": 3050
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.382782475019216,
|
| 441 |
+
"grad_norm": 5.065864086151123,
|
| 442 |
+
"learning_rate": 1.6102355808285945e-06,
|
| 443 |
+
"loss": 0.6939889526367188,
|
| 444 |
+
"step": 3100
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 2.4212144504227515,
|
| 448 |
+
"grad_norm": 4.460322380065918,
|
| 449 |
+
"learning_rate": 1.6021121039805036e-06,
|
| 450 |
+
"loss": 0.7032317352294922,
|
| 451 |
+
"step": 3150
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 2.4596464258262873,
|
| 455 |
+
"grad_norm": 5.0068817138671875,
|
| 456 |
+
"learning_rate": 1.5939886271324125e-06,
|
| 457 |
+
"loss": 0.7011449432373047,
|
| 458 |
+
"step": 3200
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.498078401229823,
|
| 462 |
+
"grad_norm": 5.634948253631592,
|
| 463 |
+
"learning_rate": 1.5858651502843216e-06,
|
| 464 |
+
"loss": 0.686719970703125,
|
| 465 |
+
"step": 3250
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 2.536510376633359,
|
| 469 |
+
"grad_norm": 4.712681293487549,
|
| 470 |
+
"learning_rate": 1.5777416734362305e-06,
|
| 471 |
+
"loss": 0.7069981384277344,
|
| 472 |
+
"step": 3300
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 2.574942352036895,
|
| 476 |
+
"grad_norm": 4.587466716766357,
|
| 477 |
+
"learning_rate": 1.5696181965881396e-06,
|
| 478 |
+
"loss": 0.7052438354492188,
|
| 479 |
+
"step": 3350
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 2.6133743274404306,
|
| 483 |
+
"grad_norm": 4.1376752853393555,
|
| 484 |
+
"learning_rate": 1.5614947197400487e-06,
|
| 485 |
+
"loss": 0.713697509765625,
|
| 486 |
+
"step": 3400
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 2.6518063028439665,
|
| 490 |
+
"grad_norm": 4.658714294433594,
|
| 491 |
+
"learning_rate": 1.5533712428919576e-06,
|
| 492 |
+
"loss": 0.697322998046875,
|
| 493 |
+
"step": 3450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 2.690238278247502,
|
| 497 |
+
"grad_norm": 4.7559685707092285,
|
| 498 |
+
"learning_rate": 1.5452477660438667e-06,
|
| 499 |
+
"loss": 0.7039759826660156,
|
| 500 |
+
"step": 3500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.7286702536510377,
|
| 504 |
+
"grad_norm": 4.555715560913086,
|
| 505 |
+
"learning_rate": 1.5371242891957756e-06,
|
| 506 |
+
"loss": 0.7013539886474609,
|
| 507 |
+
"step": 3550
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.7671022290545735,
|
| 511 |
+
"grad_norm": 4.326533317565918,
|
| 512 |
+
"learning_rate": 1.5290008123476847e-06,
|
| 513 |
+
"loss": 0.7223477172851562,
|
| 514 |
+
"step": 3600
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 2.8055342044581093,
|
| 518 |
+
"grad_norm": 4.80781364440918,
|
| 519 |
+
"learning_rate": 1.5208773354995938e-06,
|
| 520 |
+
"loss": 0.7217105102539062,
|
| 521 |
+
"step": 3650
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.8439661798616447,
|
| 525 |
+
"grad_norm": 5.230576992034912,
|
| 526 |
+
"learning_rate": 1.5127538586515027e-06,
|
| 527 |
+
"loss": 0.684557113647461,
|
| 528 |
+
"step": 3700
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 2.8823981552651805,
|
| 532 |
+
"grad_norm": 4.122222423553467,
|
| 533 |
+
"learning_rate": 1.5046303818034118e-06,
|
| 534 |
+
"loss": 0.6899368286132812,
|
| 535 |
+
"step": 3750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 2.9208301306687163,
|
| 539 |
+
"grad_norm": 4.564855575561523,
|
| 540 |
+
"learning_rate": 1.4965069049553206e-06,
|
| 541 |
+
"loss": 0.6966854858398438,
|
| 542 |
+
"step": 3800
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.959262106072252,
|
| 546 |
+
"grad_norm": 4.603681564331055,
|
| 547 |
+
"learning_rate": 1.4883834281072298e-06,
|
| 548 |
+
"loss": 0.725296630859375,
|
| 549 |
+
"step": 3850
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.997694081475788,
|
| 553 |
+
"grad_norm": 5.091521739959717,
|
| 554 |
+
"learning_rate": 1.4802599512591389e-06,
|
| 555 |
+
"loss": 0.6950398254394531,
|
| 556 |
+
"step": 3900
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 3.036126056879324,
|
| 560 |
+
"grad_norm": 4.005555629730225,
|
| 561 |
+
"learning_rate": 1.4721364744110477e-06,
|
| 562 |
+
"loss": 0.6512847900390625,
|
| 563 |
+
"step": 3950
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.074558032282859,
|
| 567 |
+
"grad_norm": 4.541317462921143,
|
| 568 |
+
"learning_rate": 1.4640129975629568e-06,
|
| 569 |
+
"loss": 0.647972183227539,
|
| 570 |
+
"step": 4000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 3.112990007686395,
|
| 574 |
+
"grad_norm": 4.654826641082764,
|
| 575 |
+
"learning_rate": 1.4558895207148657e-06,
|
| 576 |
+
"loss": 0.639323959350586,
|
| 577 |
+
"step": 4050
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 3.151421983089931,
|
| 581 |
+
"grad_norm": 4.654621124267578,
|
| 582 |
+
"learning_rate": 1.4477660438667748e-06,
|
| 583 |
+
"loss": 0.6383160018920898,
|
| 584 |
+
"step": 4100
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.1898539584934666,
|
| 588 |
+
"grad_norm": 4.840855121612549,
|
| 589 |
+
"learning_rate": 1.439642567018684e-06,
|
| 590 |
+
"loss": 0.6344610977172852,
|
| 591 |
+
"step": 4150
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 3.2282859338970025,
|
| 595 |
+
"grad_norm": 4.778560638427734,
|
| 596 |
+
"learning_rate": 1.4315190901705928e-06,
|
| 597 |
+
"loss": 0.6362688446044922,
|
| 598 |
+
"step": 4200
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 3.266717909300538,
|
| 602 |
+
"grad_norm": 4.4158477783203125,
|
| 603 |
+
"learning_rate": 1.423395613322502e-06,
|
| 604 |
+
"loss": 0.6474575805664062,
|
| 605 |
+
"step": 4250
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.3051498847040737,
|
| 609 |
+
"grad_norm": 4.960495471954346,
|
| 610 |
+
"learning_rate": 1.4152721364744108e-06,
|
| 611 |
+
"loss": 0.6327989959716797,
|
| 612 |
+
"step": 4300
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 3.3435818601076095,
|
| 616 |
+
"grad_norm": 4.7191572189331055,
|
| 617 |
+
"learning_rate": 1.40714865962632e-06,
|
| 618 |
+
"loss": 0.6356640243530274,
|
| 619 |
+
"step": 4350
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 3.3820138355111453,
|
| 623 |
+
"grad_norm": 4.39054536819458,
|
| 624 |
+
"learning_rate": 1.399025182778229e-06,
|
| 625 |
+
"loss": 0.6347926712036133,
|
| 626 |
+
"step": 4400
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.420445810914681,
|
| 630 |
+
"grad_norm": 4.320638179779053,
|
| 631 |
+
"learning_rate": 1.390901705930138e-06,
|
| 632 |
+
"loss": 0.6347624206542969,
|
| 633 |
+
"step": 4450
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 3.458877786318217,
|
| 637 |
+
"grad_norm": 4.266496181488037,
|
| 638 |
+
"learning_rate": 1.382778229082047e-06,
|
| 639 |
+
"loss": 0.6425847625732422,
|
| 640 |
+
"step": 4500
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 3.4973097617217523,
|
| 644 |
+
"grad_norm": 4.386152744293213,
|
| 645 |
+
"learning_rate": 1.374654752233956e-06,
|
| 646 |
+
"loss": 0.6452618408203125,
|
| 647 |
+
"step": 4550
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 3.535741737125288,
|
| 651 |
+
"grad_norm": 4.828922271728516,
|
| 652 |
+
"learning_rate": 1.366531275385865e-06,
|
| 653 |
+
"loss": 0.6425672912597656,
|
| 654 |
+
"step": 4600
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 3.574173712528824,
|
| 658 |
+
"grad_norm": 4.312996864318848,
|
| 659 |
+
"learning_rate": 1.3584077985377741e-06,
|
| 660 |
+
"loss": 0.6343619918823242,
|
| 661 |
+
"step": 4650
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 3.61260568793236,
|
| 665 |
+
"grad_norm": 5.0930023193359375,
|
| 666 |
+
"learning_rate": 1.350284321689683e-06,
|
| 667 |
+
"loss": 0.6599625396728516,
|
| 668 |
+
"step": 4700
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.651037663335895,
|
| 672 |
+
"grad_norm": 5.203927040100098,
|
| 673 |
+
"learning_rate": 1.3421608448415921e-06,
|
| 674 |
+
"loss": 0.6561915588378906,
|
| 675 |
+
"step": 4750
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 3.689469638739431,
|
| 679 |
+
"grad_norm": 5.211730480194092,
|
| 680 |
+
"learning_rate": 1.334037367993501e-06,
|
| 681 |
+
"loss": 0.6477365875244141,
|
| 682 |
+
"step": 4800
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 3.727901614142967,
|
| 686 |
+
"grad_norm": 4.91708517074585,
|
| 687 |
+
"learning_rate": 1.3259138911454101e-06,
|
| 688 |
+
"loss": 0.6376350402832032,
|
| 689 |
+
"step": 4850
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 3.7663335895465027,
|
| 693 |
+
"grad_norm": 4.336350917816162,
|
| 694 |
+
"learning_rate": 1.3177904142973192e-06,
|
| 695 |
+
"loss": 0.6311678314208984,
|
| 696 |
+
"step": 4900
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 3.8047655649500385,
|
| 700 |
+
"grad_norm": 4.677170276641846,
|
| 701 |
+
"learning_rate": 1.3096669374492281e-06,
|
| 702 |
+
"loss": 0.6451933288574219,
|
| 703 |
+
"step": 4950
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 3.8431975403535743,
|
| 707 |
+
"grad_norm": 4.86087703704834,
|
| 708 |
+
"learning_rate": 1.3015434606011372e-06,
|
| 709 |
+
"loss": 0.6480915069580078,
|
| 710 |
+
"step": 5000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 3.88162951575711,
|
| 714 |
+
"grad_norm": 5.272778034210205,
|
| 715 |
+
"learning_rate": 1.2934199837530461e-06,
|
| 716 |
+
"loss": 0.6448596954345703,
|
| 717 |
+
"step": 5050
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 3.9200614911606455,
|
| 721 |
+
"grad_norm": 4.808904647827148,
|
| 722 |
+
"learning_rate": 1.2852965069049552e-06,
|
| 723 |
+
"loss": 0.6556406402587891,
|
| 724 |
+
"step": 5100
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 3.9584934665641813,
|
| 728 |
+
"grad_norm": 5.201045036315918,
|
| 729 |
+
"learning_rate": 1.2771730300568643e-06,
|
| 730 |
+
"loss": 0.6454353332519531,
|
| 731 |
+
"step": 5150
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 3.996925441967717,
|
| 735 |
+
"grad_norm": 4.455353736877441,
|
| 736 |
+
"learning_rate": 1.2690495532087732e-06,
|
| 737 |
+
"loss": 0.6377482223510742,
|
| 738 |
+
"step": 5200
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 4.0353574173712525,
|
| 742 |
+
"grad_norm": 4.536952018737793,
|
| 743 |
+
"learning_rate": 1.2609260763606823e-06,
|
| 744 |
+
"loss": 0.5840181732177734,
|
| 745 |
+
"step": 5250
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.073789392774788,
|
| 749 |
+
"grad_norm": 4.572969913482666,
|
| 750 |
+
"learning_rate": 1.2528025995125912e-06,
|
| 751 |
+
"loss": 0.5863291931152343,
|
| 752 |
+
"step": 5300
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.112221368178324,
|
| 756 |
+
"grad_norm": 4.898574352264404,
|
| 757 |
+
"learning_rate": 1.2446791226645003e-06,
|
| 758 |
+
"loss": 0.5981378936767578,
|
| 759 |
+
"step": 5350
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.15065334358186,
|
| 763 |
+
"grad_norm": 4.613395690917969,
|
| 764 |
+
"learning_rate": 1.2365556458164094e-06,
|
| 765 |
+
"loss": 0.5910309600830078,
|
| 766 |
+
"step": 5400
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 4.189085318985396,
|
| 770 |
+
"grad_norm": 5.133533954620361,
|
| 771 |
+
"learning_rate": 1.2284321689683183e-06,
|
| 772 |
+
"loss": 0.5885054397583008,
|
| 773 |
+
"step": 5450
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 4.227517294388932,
|
| 777 |
+
"grad_norm": 4.249717712402344,
|
| 778 |
+
"learning_rate": 1.2203086921202274e-06,
|
| 779 |
+
"loss": 0.5926345062255859,
|
| 780 |
+
"step": 5500
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 4.2659492697924675,
|
| 784 |
+
"grad_norm": 5.025299549102783,
|
| 785 |
+
"learning_rate": 1.2121852152721363e-06,
|
| 786 |
+
"loss": 0.5963865661621094,
|
| 787 |
+
"step": 5550
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 4.304381245196003,
|
| 791 |
+
"grad_norm": 4.258217811584473,
|
| 792 |
+
"learning_rate": 1.2040617384240454e-06,
|
| 793 |
+
"loss": 0.5923780059814453,
|
| 794 |
+
"step": 5600
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 4.342813220599539,
|
| 798 |
+
"grad_norm": 4.50624418258667,
|
| 799 |
+
"learning_rate": 1.1959382615759543e-06,
|
| 800 |
+
"loss": 0.5817143249511719,
|
| 801 |
+
"step": 5650
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 4.381245196003075,
|
| 805 |
+
"grad_norm": 4.399689197540283,
|
| 806 |
+
"learning_rate": 1.1878147847278634e-06,
|
| 807 |
+
"loss": 0.5849544143676758,
|
| 808 |
+
"step": 5700
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 4.41967717140661,
|
| 812 |
+
"grad_norm": 4.704356670379639,
|
| 813 |
+
"learning_rate": 1.1796913078797725e-06,
|
| 814 |
+
"loss": 0.588057746887207,
|
| 815 |
+
"step": 5750
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 4.458109146810146,
|
| 819 |
+
"grad_norm": 4.911951541900635,
|
| 820 |
+
"learning_rate": 1.1715678310316814e-06,
|
| 821 |
+
"loss": 0.5991886138916016,
|
| 822 |
+
"step": 5800
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 4.4965411222136815,
|
| 826 |
+
"grad_norm": 4.01762580871582,
|
| 827 |
+
"learning_rate": 1.1634443541835905e-06,
|
| 828 |
+
"loss": 0.5949765777587891,
|
| 829 |
+
"step": 5850
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.534973097617217,
|
| 833 |
+
"grad_norm": 4.946042060852051,
|
| 834 |
+
"learning_rate": 1.1553208773354994e-06,
|
| 835 |
+
"loss": 0.5917446899414063,
|
| 836 |
+
"step": 5900
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.573405073020753,
|
| 840 |
+
"grad_norm": 4.350009441375732,
|
| 841 |
+
"learning_rate": 1.1471974004874085e-06,
|
| 842 |
+
"loss": 0.5994552230834961,
|
| 843 |
+
"step": 5950
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.611837048424289,
|
| 847 |
+
"grad_norm": 4.679495334625244,
|
| 848 |
+
"learning_rate": 1.1390739236393176e-06,
|
| 849 |
+
"loss": 0.5957079315185547,
|
| 850 |
+
"step": 6000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.650269023827825,
|
| 854 |
+
"grad_norm": 5.4843339920043945,
|
| 855 |
+
"learning_rate": 1.1309504467912265e-06,
|
| 856 |
+
"loss": 0.591589469909668,
|
| 857 |
+
"step": 6050
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.688700999231361,
|
| 861 |
+
"grad_norm": 4.567113399505615,
|
| 862 |
+
"learning_rate": 1.1228269699431356e-06,
|
| 863 |
+
"loss": 0.5838856506347656,
|
| 864 |
+
"step": 6100
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.7271329746348965,
|
| 868 |
+
"grad_norm": 5.122208595275879,
|
| 869 |
+
"learning_rate": 1.1147034930950445e-06,
|
| 870 |
+
"loss": 0.6134935760498047,
|
| 871 |
+
"step": 6150
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.765564950038432,
|
| 875 |
+
"grad_norm": 4.586711406707764,
|
| 876 |
+
"learning_rate": 1.1065800162469536e-06,
|
| 877 |
+
"loss": 0.5973595428466797,
|
| 878 |
+
"step": 6200
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.803996925441968,
|
| 882 |
+
"grad_norm": 4.636633396148682,
|
| 883 |
+
"learning_rate": 1.0984565393988627e-06,
|
| 884 |
+
"loss": 0.5989643096923828,
|
| 885 |
+
"step": 6250
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.842428900845503,
|
| 889 |
+
"grad_norm": 4.6541523933410645,
|
| 890 |
+
"learning_rate": 1.0903330625507716e-06,
|
| 891 |
+
"loss": 0.5864573669433594,
|
| 892 |
+
"step": 6300
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.880860876249039,
|
| 896 |
+
"grad_norm": 4.1167826652526855,
|
| 897 |
+
"learning_rate": 1.0822095857026807e-06,
|
| 898 |
+
"loss": 0.5802556610107422,
|
| 899 |
+
"step": 6350
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.919292851652575,
|
| 903 |
+
"grad_norm": 4.347712516784668,
|
| 904 |
+
"learning_rate": 1.0740861088545896e-06,
|
| 905 |
+
"loss": 0.5990000915527344,
|
| 906 |
+
"step": 6400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.9577248270561105,
|
| 910 |
+
"grad_norm": 4.584090232849121,
|
| 911 |
+
"learning_rate": 1.0659626320064987e-06,
|
| 912 |
+
"loss": 0.6011124420166015,
|
| 913 |
+
"step": 6450
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.996156802459646,
|
| 917 |
+
"grad_norm": 4.22511100769043,
|
| 918 |
+
"learning_rate": 1.0578391551584078e-06,
|
| 919 |
+
"loss": 0.597344970703125,
|
| 920 |
+
"step": 6500
|
| 921 |
+
}
|
| 922 |
+
],
|
| 923 |
+
"logging_steps": 50,
|
| 924 |
+
"max_steps": 13010,
|
| 925 |
+
"num_input_tokens_seen": 0,
|
| 926 |
+
"num_train_epochs": 10,
|
| 927 |
+
"save_steps": 500,
|
| 928 |
+
"stateful_callbacks": {
|
| 929 |
+
"TrainerControl": {
|
| 930 |
+
"args": {
|
| 931 |
+
"should_epoch_stop": false,
|
| 932 |
+
"should_evaluate": false,
|
| 933 |
+
"should_log": false,
|
| 934 |
+
"should_save": true,
|
| 935 |
+
"should_training_stop": false
|
| 936 |
+
},
|
| 937 |
+
"attributes": {}
|
| 938 |
+
}
|
| 939 |
+
},
|
| 940 |
+
"total_flos": 8.497964789984448e+20,
|
| 941 |
+
"train_batch_size": 16,
|
| 942 |
+
"trial_name": null,
|
| 943 |
+
"trial_params": null
|
| 944 |
+
}
|
checkpoint-6505/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f1eba0b8d773b390681a5af8baaa03011d99a516e1f9d874bad4876e2fe90ee
|
| 3 |
+
size 5329
|