Upload folder using huggingface_hub
Browse files- config.json +233 -0
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1089 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,233 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "facebook/wav2vec2-xls-r-300m",
|
| 3 |
+
"activation_dropout": 0.0,
|
| 4 |
+
"adapter_attn_dim": null,
|
| 5 |
+
"adapter_kernel_size": 3,
|
| 6 |
+
"adapter_stride": 2,
|
| 7 |
+
"add_adapter": false,
|
| 8 |
+
"apply_spec_augment": true,
|
| 9 |
+
"architectures": [
|
| 10 |
+
"Wav2Vec2ForSequenceClassification"
|
| 11 |
+
],
|
| 12 |
+
"attention_dropout": 0.1,
|
| 13 |
+
"bos_token_id": 1,
|
| 14 |
+
"classifier_proj_size": 256,
|
| 15 |
+
"codevector_dim": 768,
|
| 16 |
+
"contrastive_logits_temperature": 0.1,
|
| 17 |
+
"conv_bias": true,
|
| 18 |
+
"conv_dim": [
|
| 19 |
+
512,
|
| 20 |
+
512,
|
| 21 |
+
512,
|
| 22 |
+
512,
|
| 23 |
+
512,
|
| 24 |
+
512,
|
| 25 |
+
512
|
| 26 |
+
],
|
| 27 |
+
"conv_kernel": [
|
| 28 |
+
10,
|
| 29 |
+
3,
|
| 30 |
+
3,
|
| 31 |
+
3,
|
| 32 |
+
3,
|
| 33 |
+
2,
|
| 34 |
+
2
|
| 35 |
+
],
|
| 36 |
+
"conv_stride": [
|
| 37 |
+
5,
|
| 38 |
+
2,
|
| 39 |
+
2,
|
| 40 |
+
2,
|
| 41 |
+
2,
|
| 42 |
+
2,
|
| 43 |
+
2
|
| 44 |
+
],
|
| 45 |
+
"ctc_loss_reduction": "sum",
|
| 46 |
+
"ctc_zero_infinity": false,
|
| 47 |
+
"diversity_loss_weight": 0.1,
|
| 48 |
+
"do_stable_layer_norm": true,
|
| 49 |
+
"eos_token_id": 2,
|
| 50 |
+
"feat_extract_activation": "gelu",
|
| 51 |
+
"feat_extract_dropout": 0.0,
|
| 52 |
+
"feat_extract_norm": "layer",
|
| 53 |
+
"feat_proj_dropout": 0.1,
|
| 54 |
+
"feat_quantizer_dropout": 0.0,
|
| 55 |
+
"final_dropout": 0.0,
|
| 56 |
+
"gradient_checkpointing": false,
|
| 57 |
+
"hidden_act": "gelu",
|
| 58 |
+
"hidden_dropout": 0.1,
|
| 59 |
+
"hidden_size": 1024,
|
| 60 |
+
"id2label": {
|
| 61 |
+
"0": "iot_hue_lighton",
|
| 62 |
+
"1": "takeaway_order",
|
| 63 |
+
"10": "iot_wemo_off",
|
| 64 |
+
"11": "qa_maths",
|
| 65 |
+
"12": "iot_wemo_on",
|
| 66 |
+
"13": "iot_coffee",
|
| 67 |
+
"14": "email_query",
|
| 68 |
+
"15": "alarm_set",
|
| 69 |
+
"16": "lists_query",
|
| 70 |
+
"17": "audio_volume_mute",
|
| 71 |
+
"18": "music_query",
|
| 72 |
+
"19": "transport_query",
|
| 73 |
+
"2": "social_query",
|
| 74 |
+
"20": "social_post",
|
| 75 |
+
"21": "qa_stock",
|
| 76 |
+
"22": "qa_definition",
|
| 77 |
+
"23": "cooking_recipe",
|
| 78 |
+
"24": "qa_currency",
|
| 79 |
+
"25": "email_querycontact",
|
| 80 |
+
"26": "iot_hue_lightoff",
|
| 81 |
+
"27": "transport_taxi",
|
| 82 |
+
"28": "recommendation_events",
|
| 83 |
+
"29": "email_addcontact",
|
| 84 |
+
"3": "qa_factoid",
|
| 85 |
+
"30": "recommendation_locations",
|
| 86 |
+
"31": "email_sendemail",
|
| 87 |
+
"32": "play_podcasts",
|
| 88 |
+
"33": "cooking_query",
|
| 89 |
+
"34": "audio_volume_other",
|
| 90 |
+
"35": "play_music",
|
| 91 |
+
"36": "calendar_remove",
|
| 92 |
+
"37": "datetime_query",
|
| 93 |
+
"38": "weather_query",
|
| 94 |
+
"39": "recommendation_movies",
|
| 95 |
+
"4": "takeaway_query",
|
| 96 |
+
"40": "alarm_remove",
|
| 97 |
+
"41": "audio_volume_down",
|
| 98 |
+
"42": "music_settings",
|
| 99 |
+
"43": "music_dislikeness",
|
| 100 |
+
"44": "iot_cleaning",
|
| 101 |
+
"45": "iot_hue_lightup",
|
| 102 |
+
"46": "audio_volume_up",
|
| 103 |
+
"47": "general_quirky",
|
| 104 |
+
"48": "transport_ticket",
|
| 105 |
+
"49": "iot_hue_lightdim",
|
| 106 |
+
"5": "lists_remove",
|
| 107 |
+
"50": "general_greet",
|
| 108 |
+
"51": "datetime_convert",
|
| 109 |
+
"52": "calendar_query",
|
| 110 |
+
"53": "play_radio",
|
| 111 |
+
"54": "alarm_query",
|
| 112 |
+
"55": "transport_traffic",
|
| 113 |
+
"56": "calendar_set",
|
| 114 |
+
"57": "lists_createoradd",
|
| 115 |
+
"58": "play_game",
|
| 116 |
+
"59": "music_likeness",
|
| 117 |
+
"6": "iot_hue_lightchange",
|
| 118 |
+
"7": "play_audiobook",
|
| 119 |
+
"8": "general_joke",
|
| 120 |
+
"9": "news_query"
|
| 121 |
+
},
|
| 122 |
+
"initializer_range": 0.02,
|
| 123 |
+
"intermediate_size": 4096,
|
| 124 |
+
"label2id": {
|
| 125 |
+
"alarm_query": "54",
|
| 126 |
+
"alarm_remove": "40",
|
| 127 |
+
"alarm_set": "15",
|
| 128 |
+
"audio_volume_down": "41",
|
| 129 |
+
"audio_volume_mute": "17",
|
| 130 |
+
"audio_volume_other": "34",
|
| 131 |
+
"audio_volume_up": "46",
|
| 132 |
+
"calendar_query": "52",
|
| 133 |
+
"calendar_remove": "36",
|
| 134 |
+
"calendar_set": "56",
|
| 135 |
+
"cooking_query": "33",
|
| 136 |
+
"cooking_recipe": "23",
|
| 137 |
+
"datetime_convert": "51",
|
| 138 |
+
"datetime_query": "37",
|
| 139 |
+
"email_addcontact": "29",
|
| 140 |
+
"email_query": "14",
|
| 141 |
+
"email_querycontact": "25",
|
| 142 |
+
"email_sendemail": "31",
|
| 143 |
+
"general_greet": "50",
|
| 144 |
+
"general_joke": "8",
|
| 145 |
+
"general_quirky": "47",
|
| 146 |
+
"iot_cleaning": "44",
|
| 147 |
+
"iot_coffee": "13",
|
| 148 |
+
"iot_hue_lightchange": "6",
|
| 149 |
+
"iot_hue_lightdim": "49",
|
| 150 |
+
"iot_hue_lightoff": "26",
|
| 151 |
+
"iot_hue_lighton": "0",
|
| 152 |
+
"iot_hue_lightup": "45",
|
| 153 |
+
"iot_wemo_off": "10",
|
| 154 |
+
"iot_wemo_on": "12",
|
| 155 |
+
"lists_createoradd": "57",
|
| 156 |
+
"lists_query": "16",
|
| 157 |
+
"lists_remove": "5",
|
| 158 |
+
"music_dislikeness": "43",
|
| 159 |
+
"music_likeness": "59",
|
| 160 |
+
"music_query": "18",
|
| 161 |
+
"music_settings": "42",
|
| 162 |
+
"news_query": "9",
|
| 163 |
+
"play_audiobook": "7",
|
| 164 |
+
"play_game": "58",
|
| 165 |
+
"play_music": "35",
|
| 166 |
+
"play_podcasts": "32",
|
| 167 |
+
"play_radio": "53",
|
| 168 |
+
"qa_currency": "24",
|
| 169 |
+
"qa_definition": "22",
|
| 170 |
+
"qa_factoid": "3",
|
| 171 |
+
"qa_maths": "11",
|
| 172 |
+
"qa_stock": "21",
|
| 173 |
+
"recommendation_events": "28",
|
| 174 |
+
"recommendation_locations": "30",
|
| 175 |
+
"recommendation_movies": "39",
|
| 176 |
+
"social_post": "20",
|
| 177 |
+
"social_query": "2",
|
| 178 |
+
"takeaway_order": "1",
|
| 179 |
+
"takeaway_query": "4",
|
| 180 |
+
"transport_query": "19",
|
| 181 |
+
"transport_taxi": "27",
|
| 182 |
+
"transport_ticket": "48",
|
| 183 |
+
"transport_traffic": "55",
|
| 184 |
+
"weather_query": "38"
|
| 185 |
+
},
|
| 186 |
+
"layer_norm_eps": 1e-05,
|
| 187 |
+
"layerdrop": 0.1,
|
| 188 |
+
"mask_feature_length": 10,
|
| 189 |
+
"mask_feature_min_masks": 0,
|
| 190 |
+
"mask_feature_prob": 0.0,
|
| 191 |
+
"mask_time_length": 10,
|
| 192 |
+
"mask_time_min_masks": 2,
|
| 193 |
+
"mask_time_prob": 0.075,
|
| 194 |
+
"model_type": "wav2vec2",
|
| 195 |
+
"num_adapter_layers": 3,
|
| 196 |
+
"num_attention_heads": 16,
|
| 197 |
+
"num_codevector_groups": 2,
|
| 198 |
+
"num_codevectors_per_group": 320,
|
| 199 |
+
"num_conv_pos_embedding_groups": 16,
|
| 200 |
+
"num_conv_pos_embeddings": 128,
|
| 201 |
+
"num_feat_extract_layers": 7,
|
| 202 |
+
"num_hidden_layers": 24,
|
| 203 |
+
"num_negatives": 100,
|
| 204 |
+
"output_hidden_size": 1024,
|
| 205 |
+
"pad_token_id": 0,
|
| 206 |
+
"proj_codevector_dim": 768,
|
| 207 |
+
"tdnn_dilation": [
|
| 208 |
+
1,
|
| 209 |
+
2,
|
| 210 |
+
3,
|
| 211 |
+
1,
|
| 212 |
+
1
|
| 213 |
+
],
|
| 214 |
+
"tdnn_dim": [
|
| 215 |
+
512,
|
| 216 |
+
512,
|
| 217 |
+
512,
|
| 218 |
+
512,
|
| 219 |
+
1500
|
| 220 |
+
],
|
| 221 |
+
"tdnn_kernel": [
|
| 222 |
+
5,
|
| 223 |
+
3,
|
| 224 |
+
3,
|
| 225 |
+
1,
|
| 226 |
+
1
|
| 227 |
+
],
|
| 228 |
+
"torch_dtype": "float32",
|
| 229 |
+
"transformers_version": "4.45.2",
|
| 230 |
+
"use_weighted_layer_sum": false,
|
| 231 |
+
"vocab_size": 32,
|
| 232 |
+
"xvector_output_dim": 512
|
| 233 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a65bb9e103306975927fad1e19fb21ac4d45f61ea79f2983c7965daef7752700
|
| 3 |
+
size 1262918936
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e2db683850fcdb20d41ea16d7f0b6c2ff5ece52f4d78cb5b566c139a6d1a06d
|
| 3 |
+
size 2526088912
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1ec380b95237a53030c7415c534e37a484a51a38c10ad193d7ce48d2a588974
|
| 3 |
+
size 14308
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81c3cab4775743cdbcd968af79fd32c9628832c84d1ee5982b7ec308e6969676
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1089 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.691096901131333,
|
| 3 |
+
"best_model_checkpoint": "de-DE/wav2vec2-xls-r-300m/checkpoint-5040",
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 5400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.2777777777777778,
|
| 13 |
+
"grad_norm": 1.4890042543411255,
|
| 14 |
+
"learning_rate": 9.074074074074075e-06,
|
| 15 |
+
"loss": 4.091,
|
| 16 |
+
"step": 50
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.5555555555555556,
|
| 20 |
+
"grad_norm": 1.1139342784881592,
|
| 21 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 22 |
+
"loss": 4.0628,
|
| 23 |
+
"step": 100
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.8333333333333334,
|
| 27 |
+
"grad_norm": 0.7727408409118652,
|
| 28 |
+
"learning_rate": 2.7592592592592594e-05,
|
| 29 |
+
"loss": 3.9799,
|
| 30 |
+
"step": 150
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.0,
|
| 34 |
+
"eval_accuracy": 0.06443679291687161,
|
| 35 |
+
"eval_f1": 0.00205206930041668,
|
| 36 |
+
"eval_loss": 3.8087375164031982,
|
| 37 |
+
"eval_runtime": 21.0392,
|
| 38 |
+
"eval_samples_per_second": 96.629,
|
| 39 |
+
"eval_steps_per_second": 1.521,
|
| 40 |
+
"step": 180
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 1.1111111111111112,
|
| 44 |
+
"grad_norm": 1.1827894449234009,
|
| 45 |
+
"learning_rate": 3.685185185185185e-05,
|
| 46 |
+
"loss": 3.8573,
|
| 47 |
+
"step": 200
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 1.3888888888888888,
|
| 51 |
+
"grad_norm": 1.3509910106658936,
|
| 52 |
+
"learning_rate": 4.6111111111111115e-05,
|
| 53 |
+
"loss": 3.7732,
|
| 54 |
+
"step": 250
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 1.6666666666666665,
|
| 58 |
+
"grad_norm": 1.5033173561096191,
|
| 59 |
+
"learning_rate": 5.537037037037037e-05,
|
| 60 |
+
"loss": 3.7526,
|
| 61 |
+
"step": 300
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 1.9444444444444444,
|
| 65 |
+
"grad_norm": 1.2770198583602905,
|
| 66 |
+
"learning_rate": 6.462962962962962e-05,
|
| 67 |
+
"loss": 3.7773,
|
| 68 |
+
"step": 350
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 2.0,
|
| 72 |
+
"eval_accuracy": 0.06443679291687161,
|
| 73 |
+
"eval_f1": 0.00205206930041668,
|
| 74 |
+
"eval_loss": 3.736783504486084,
|
| 75 |
+
"eval_runtime": 21.1812,
|
| 76 |
+
"eval_samples_per_second": 95.982,
|
| 77 |
+
"eval_steps_per_second": 1.511,
|
| 78 |
+
"step": 360
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 2.2222222222222223,
|
| 82 |
+
"grad_norm": 1.2693891525268555,
|
| 83 |
+
"learning_rate": 7.38888888888889e-05,
|
| 84 |
+
"loss": 3.7473,
|
| 85 |
+
"step": 400
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 2.5,
|
| 89 |
+
"grad_norm": 1.4083213806152344,
|
| 90 |
+
"learning_rate": 8.314814814814815e-05,
|
| 91 |
+
"loss": 3.7626,
|
| 92 |
+
"step": 450
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 2.7777777777777777,
|
| 96 |
+
"grad_norm": 1.625786542892456,
|
| 97 |
+
"learning_rate": 9.240740740740741e-05,
|
| 98 |
+
"loss": 3.7632,
|
| 99 |
+
"step": 500
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 3.0,
|
| 103 |
+
"eval_accuracy": 0.06443679291687161,
|
| 104 |
+
"eval_f1": 0.00205206930041668,
|
| 105 |
+
"eval_loss": 3.734524965286255,
|
| 106 |
+
"eval_runtime": 21.1901,
|
| 107 |
+
"eval_samples_per_second": 95.941,
|
| 108 |
+
"eval_steps_per_second": 1.51,
|
| 109 |
+
"step": 540
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 3.0555555555555554,
|
| 113 |
+
"grad_norm": 1.2629117965698242,
|
| 114 |
+
"learning_rate": 9.981481481481481e-05,
|
| 115 |
+
"loss": 3.7604,
|
| 116 |
+
"step": 550
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 3.3333333333333335,
|
| 120 |
+
"grad_norm": 1.164554476737976,
|
| 121 |
+
"learning_rate": 9.878600823045268e-05,
|
| 122 |
+
"loss": 3.7564,
|
| 123 |
+
"step": 600
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 3.611111111111111,
|
| 127 |
+
"grad_norm": 0.9783994555473328,
|
| 128 |
+
"learning_rate": 9.775720164609054e-05,
|
| 129 |
+
"loss": 3.7596,
|
| 130 |
+
"step": 650
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 3.888888888888889,
|
| 134 |
+
"grad_norm": 1.4372780323028564,
|
| 135 |
+
"learning_rate": 9.672839506172841e-05,
|
| 136 |
+
"loss": 3.7706,
|
| 137 |
+
"step": 700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 4.0,
|
| 141 |
+
"eval_accuracy": 0.06443679291687161,
|
| 142 |
+
"eval_f1": 0.00205206930041668,
|
| 143 |
+
"eval_loss": 3.737985372543335,
|
| 144 |
+
"eval_runtime": 20.8407,
|
| 145 |
+
"eval_samples_per_second": 97.55,
|
| 146 |
+
"eval_steps_per_second": 1.535,
|
| 147 |
+
"step": 720
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 4.166666666666667,
|
| 151 |
+
"grad_norm": 1.2122852802276611,
|
| 152 |
+
"learning_rate": 9.569958847736626e-05,
|
| 153 |
+
"loss": 3.7516,
|
| 154 |
+
"step": 750
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 4.444444444444445,
|
| 158 |
+
"grad_norm": 1.3332332372665405,
|
| 159 |
+
"learning_rate": 9.467078189300412e-05,
|
| 160 |
+
"loss": 3.735,
|
| 161 |
+
"step": 800
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 4.722222222222222,
|
| 165 |
+
"grad_norm": 0.9196431040763855,
|
| 166 |
+
"learning_rate": 9.364197530864199e-05,
|
| 167 |
+
"loss": 3.7224,
|
| 168 |
+
"step": 850
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 5.0,
|
| 172 |
+
"grad_norm": 1.3251221179962158,
|
| 173 |
+
"learning_rate": 9.261316872427984e-05,
|
| 174 |
+
"loss": 3.7518,
|
| 175 |
+
"step": 900
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 5.0,
|
| 179 |
+
"eval_accuracy": 0.09296606000983768,
|
| 180 |
+
"eval_f1": 0.0059176939246660745,
|
| 181 |
+
"eval_loss": 3.6887905597686768,
|
| 182 |
+
"eval_runtime": 21.2519,
|
| 183 |
+
"eval_samples_per_second": 95.662,
|
| 184 |
+
"eval_steps_per_second": 1.506,
|
| 185 |
+
"step": 900
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 5.277777777777778,
|
| 189 |
+
"grad_norm": 1.470018982887268,
|
| 190 |
+
"learning_rate": 9.15843621399177e-05,
|
| 191 |
+
"loss": 3.7136,
|
| 192 |
+
"step": 950
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 5.555555555555555,
|
| 196 |
+
"grad_norm": 1.4413223266601562,
|
| 197 |
+
"learning_rate": 9.055555555555556e-05,
|
| 198 |
+
"loss": 3.6818,
|
| 199 |
+
"step": 1000
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 5.833333333333333,
|
| 203 |
+
"grad_norm": 1.2150436639785767,
|
| 204 |
+
"learning_rate": 8.952674897119342e-05,
|
| 205 |
+
"loss": 3.699,
|
| 206 |
+
"step": 1050
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 6.0,
|
| 210 |
+
"eval_accuracy": 0.09936055090998525,
|
| 211 |
+
"eval_f1": 0.0073766292774243535,
|
| 212 |
+
"eval_loss": 3.6655001640319824,
|
| 213 |
+
"eval_runtime": 20.9819,
|
| 214 |
+
"eval_samples_per_second": 96.893,
|
| 215 |
+
"eval_steps_per_second": 1.525,
|
| 216 |
+
"step": 1080
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 6.111111111111111,
|
| 220 |
+
"grad_norm": 1.4290884733200073,
|
| 221 |
+
"learning_rate": 8.849794238683128e-05,
|
| 222 |
+
"loss": 3.6581,
|
| 223 |
+
"step": 1100
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 6.388888888888889,
|
| 227 |
+
"grad_norm": 1.6972917318344116,
|
| 228 |
+
"learning_rate": 8.746913580246914e-05,
|
| 229 |
+
"loss": 3.6401,
|
| 230 |
+
"step": 1150
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 6.666666666666667,
|
| 234 |
+
"grad_norm": 1.3264811038970947,
|
| 235 |
+
"learning_rate": 8.644032921810699e-05,
|
| 236 |
+
"loss": 3.6843,
|
| 237 |
+
"step": 1200
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 6.944444444444445,
|
| 241 |
+
"grad_norm": 1.6939361095428467,
|
| 242 |
+
"learning_rate": 8.541152263374486e-05,
|
| 243 |
+
"loss": 3.6017,
|
| 244 |
+
"step": 1250
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 7.0,
|
| 248 |
+
"eval_accuracy": 0.11116576487948844,
|
| 249 |
+
"eval_f1": 0.007963237845435386,
|
| 250 |
+
"eval_loss": 3.613088846206665,
|
| 251 |
+
"eval_runtime": 21.1422,
|
| 252 |
+
"eval_samples_per_second": 96.158,
|
| 253 |
+
"eval_steps_per_second": 1.514,
|
| 254 |
+
"step": 1260
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 7.222222222222222,
|
| 258 |
+
"grad_norm": 1.3342170715332031,
|
| 259 |
+
"learning_rate": 8.438271604938272e-05,
|
| 260 |
+
"loss": 3.5671,
|
| 261 |
+
"step": 1300
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 7.5,
|
| 265 |
+
"grad_norm": 2.0039055347442627,
|
| 266 |
+
"learning_rate": 8.335390946502059e-05,
|
| 267 |
+
"loss": 3.563,
|
| 268 |
+
"step": 1350
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 7.777777777777778,
|
| 272 |
+
"grad_norm": 1.9838131666183472,
|
| 273 |
+
"learning_rate": 8.232510288065844e-05,
|
| 274 |
+
"loss": 3.4976,
|
| 275 |
+
"step": 1400
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 8.0,
|
| 279 |
+
"eval_accuracy": 0.12297097884899164,
|
| 280 |
+
"eval_f1": 0.0207489285319609,
|
| 281 |
+
"eval_loss": 3.5292012691497803,
|
| 282 |
+
"eval_runtime": 21.1877,
|
| 283 |
+
"eval_samples_per_second": 95.952,
|
| 284 |
+
"eval_steps_per_second": 1.51,
|
| 285 |
+
"step": 1440
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 8.055555555555555,
|
| 289 |
+
"grad_norm": 1.7923851013183594,
|
| 290 |
+
"learning_rate": 8.12962962962963e-05,
|
| 291 |
+
"loss": 3.5203,
|
| 292 |
+
"step": 1450
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 8.333333333333334,
|
| 296 |
+
"grad_norm": 2.5529909133911133,
|
| 297 |
+
"learning_rate": 8.026748971193417e-05,
|
| 298 |
+
"loss": 3.4522,
|
| 299 |
+
"step": 1500
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 8.61111111111111,
|
| 303 |
+
"grad_norm": 1.8508727550506592,
|
| 304 |
+
"learning_rate": 7.925925925925926e-05,
|
| 305 |
+
"loss": 3.418,
|
| 306 |
+
"step": 1550
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 8.88888888888889,
|
| 310 |
+
"grad_norm": 2.031550407409668,
|
| 311 |
+
"learning_rate": 7.823045267489712e-05,
|
| 312 |
+
"loss": 3.379,
|
| 313 |
+
"step": 1600
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 9.0,
|
| 317 |
+
"eval_accuracy": 0.14805705853418594,
|
| 318 |
+
"eval_f1": 0.030926614730120143,
|
| 319 |
+
"eval_loss": 3.388072967529297,
|
| 320 |
+
"eval_runtime": 21.1544,
|
| 321 |
+
"eval_samples_per_second": 96.103,
|
| 322 |
+
"eval_steps_per_second": 1.513,
|
| 323 |
+
"step": 1620
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 9.166666666666666,
|
| 327 |
+
"grad_norm": 2.7599411010742188,
|
| 328 |
+
"learning_rate": 7.720164609053499e-05,
|
| 329 |
+
"loss": 3.3362,
|
| 330 |
+
"step": 1650
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 9.444444444444445,
|
| 334 |
+
"grad_norm": 3.2844510078430176,
|
| 335 |
+
"learning_rate": 7.617283950617285e-05,
|
| 336 |
+
"loss": 3.3089,
|
| 337 |
+
"step": 1700
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 9.722222222222221,
|
| 341 |
+
"grad_norm": 3.1966094970703125,
|
| 342 |
+
"learning_rate": 7.516460905349794e-05,
|
| 343 |
+
"loss": 3.2753,
|
| 344 |
+
"step": 1750
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 10.0,
|
| 348 |
+
"grad_norm": 3.316312074661255,
|
| 349 |
+
"learning_rate": 7.413580246913581e-05,
|
| 350 |
+
"loss": 3.2182,
|
| 351 |
+
"step": 1800
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 10.0,
|
| 355 |
+
"eval_accuracy": 0.19527791441219872,
|
| 356 |
+
"eval_f1": 0.04058892684496088,
|
| 357 |
+
"eval_loss": 3.1913669109344482,
|
| 358 |
+
"eval_runtime": 20.9617,
|
| 359 |
+
"eval_samples_per_second": 96.986,
|
| 360 |
+
"eval_steps_per_second": 1.527,
|
| 361 |
+
"step": 1800
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 10.277777777777779,
|
| 365 |
+
"grad_norm": 3.078585624694824,
|
| 366 |
+
"learning_rate": 7.310699588477367e-05,
|
| 367 |
+
"loss": 3.0993,
|
| 368 |
+
"step": 1850
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 10.555555555555555,
|
| 372 |
+
"grad_norm": 6.923484802246094,
|
| 373 |
+
"learning_rate": 7.207818930041152e-05,
|
| 374 |
+
"loss": 3.0807,
|
| 375 |
+
"step": 1900
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 10.833333333333334,
|
| 379 |
+
"grad_norm": 5.272449970245361,
|
| 380 |
+
"learning_rate": 7.104938271604938e-05,
|
| 381 |
+
"loss": 2.9804,
|
| 382 |
+
"step": 1950
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 11.0,
|
| 386 |
+
"eval_accuracy": 0.2267584849975406,
|
| 387 |
+
"eval_f1": 0.05879115082822113,
|
| 388 |
+
"eval_loss": 2.93943452835083,
|
| 389 |
+
"eval_runtime": 21.2521,
|
| 390 |
+
"eval_samples_per_second": 95.661,
|
| 391 |
+
"eval_steps_per_second": 1.506,
|
| 392 |
+
"step": 1980
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 11.11111111111111,
|
| 396 |
+
"grad_norm": 7.379478931427002,
|
| 397 |
+
"learning_rate": 7.002057613168725e-05,
|
| 398 |
+
"loss": 2.9518,
|
| 399 |
+
"step": 2000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 11.38888888888889,
|
| 403 |
+
"grad_norm": 4.347708225250244,
|
| 404 |
+
"learning_rate": 6.89917695473251e-05,
|
| 405 |
+
"loss": 2.8607,
|
| 406 |
+
"step": 2050
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 11.666666666666666,
|
| 410 |
+
"grad_norm": 3.146164894104004,
|
| 411 |
+
"learning_rate": 6.796296296296296e-05,
|
| 412 |
+
"loss": 2.805,
|
| 413 |
+
"step": 2100
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 11.944444444444445,
|
| 417 |
+
"grad_norm": 3.2119126319885254,
|
| 418 |
+
"learning_rate": 6.693415637860082e-05,
|
| 419 |
+
"loss": 2.7577,
|
| 420 |
+
"step": 2150
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 12.0,
|
| 424 |
+
"eval_accuracy": 0.2838170191834727,
|
| 425 |
+
"eval_f1": 0.09477911894276655,
|
| 426 |
+
"eval_loss": 2.6934072971343994,
|
| 427 |
+
"eval_runtime": 21.1897,
|
| 428 |
+
"eval_samples_per_second": 95.943,
|
| 429 |
+
"eval_steps_per_second": 1.51,
|
| 430 |
+
"step": 2160
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 12.222222222222221,
|
| 434 |
+
"grad_norm": 5.8592329025268555,
|
| 435 |
+
"learning_rate": 6.590534979423869e-05,
|
| 436 |
+
"loss": 2.6594,
|
| 437 |
+
"step": 2200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 12.5,
|
| 441 |
+
"grad_norm": 3.185109853744507,
|
| 442 |
+
"learning_rate": 6.487654320987656e-05,
|
| 443 |
+
"loss": 2.5666,
|
| 444 |
+
"step": 2250
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 12.777777777777779,
|
| 448 |
+
"grad_norm": 6.137824535369873,
|
| 449 |
+
"learning_rate": 6.384773662551441e-05,
|
| 450 |
+
"loss": 2.5161,
|
| 451 |
+
"step": 2300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 13.0,
|
| 455 |
+
"eval_accuracy": 0.3320216428922774,
|
| 456 |
+
"eval_f1": 0.13001894064725447,
|
| 457 |
+
"eval_loss": 2.445955753326416,
|
| 458 |
+
"eval_runtime": 21.1936,
|
| 459 |
+
"eval_samples_per_second": 95.925,
|
| 460 |
+
"eval_steps_per_second": 1.51,
|
| 461 |
+
"step": 2340
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 13.055555555555555,
|
| 465 |
+
"grad_norm": 3.701847791671753,
|
| 466 |
+
"learning_rate": 6.281893004115227e-05,
|
| 467 |
+
"loss": 2.408,
|
| 468 |
+
"step": 2350
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 13.333333333333334,
|
| 472 |
+
"grad_norm": 3.357701539993286,
|
| 473 |
+
"learning_rate": 6.179012345679012e-05,
|
| 474 |
+
"loss": 2.2915,
|
| 475 |
+
"step": 2400
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 13.61111111111111,
|
| 479 |
+
"grad_norm": 4.91466760635376,
|
| 480 |
+
"learning_rate": 6.0761316872427985e-05,
|
| 481 |
+
"loss": 2.2576,
|
| 482 |
+
"step": 2450
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 13.88888888888889,
|
| 486 |
+
"grad_norm": 4.737729072570801,
|
| 487 |
+
"learning_rate": 5.973251028806585e-05,
|
| 488 |
+
"loss": 2.2414,
|
| 489 |
+
"step": 2500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 14.0,
|
| 493 |
+
"eval_accuracy": 0.3890801770782095,
|
| 494 |
+
"eval_f1": 0.1854455337191276,
|
| 495 |
+
"eval_loss": 2.230682373046875,
|
| 496 |
+
"eval_runtime": 21.2287,
|
| 497 |
+
"eval_samples_per_second": 95.767,
|
| 498 |
+
"eval_steps_per_second": 1.507,
|
| 499 |
+
"step": 2520
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 14.166666666666666,
|
| 503 |
+
"grad_norm": 12.414363861083984,
|
| 504 |
+
"learning_rate": 5.8703703703703704e-05,
|
| 505 |
+
"loss": 2.125,
|
| 506 |
+
"step": 2550
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 14.444444444444445,
|
| 510 |
+
"grad_norm": 4.032175540924072,
|
| 511 |
+
"learning_rate": 5.7674897119341566e-05,
|
| 512 |
+
"loss": 2.0399,
|
| 513 |
+
"step": 2600
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 14.722222222222221,
|
| 517 |
+
"grad_norm": 5.251100540161133,
|
| 518 |
+
"learning_rate": 5.664609053497942e-05,
|
| 519 |
+
"loss": 2.023,
|
| 520 |
+
"step": 2650
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 15.0,
|
| 524 |
+
"grad_norm": 4.966403484344482,
|
| 525 |
+
"learning_rate": 5.5617283950617284e-05,
|
| 526 |
+
"loss": 1.944,
|
| 527 |
+
"step": 2700
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 15.0,
|
| 531 |
+
"eval_accuracy": 0.45204131824889326,
|
| 532 |
+
"eval_f1": 0.2558497963614674,
|
| 533 |
+
"eval_loss": 2.048044204711914,
|
| 534 |
+
"eval_runtime": 21.1364,
|
| 535 |
+
"eval_samples_per_second": 96.185,
|
| 536 |
+
"eval_steps_per_second": 1.514,
|
| 537 |
+
"step": 2700
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 15.277777777777779,
|
| 541 |
+
"grad_norm": 4.946567535400391,
|
| 542 |
+
"learning_rate": 5.458847736625514e-05,
|
| 543 |
+
"loss": 1.7941,
|
| 544 |
+
"step": 2750
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 15.555555555555555,
|
| 548 |
+
"grad_norm": 6.810694217681885,
|
| 549 |
+
"learning_rate": 5.355967078189301e-05,
|
| 550 |
+
"loss": 1.7366,
|
| 551 |
+
"step": 2800
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 15.833333333333334,
|
| 555 |
+
"grad_norm": 7.022438049316406,
|
| 556 |
+
"learning_rate": 5.253086419753087e-05,
|
| 557 |
+
"loss": 1.7764,
|
| 558 |
+
"step": 2850
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 16.0,
|
| 562 |
+
"eval_accuracy": 0.5081160846040335,
|
| 563 |
+
"eval_f1": 0.32205117403665584,
|
| 564 |
+
"eval_loss": 1.9176976680755615,
|
| 565 |
+
"eval_runtime": 21.2753,
|
| 566 |
+
"eval_samples_per_second": 95.557,
|
| 567 |
+
"eval_steps_per_second": 1.504,
|
| 568 |
+
"step": 2880
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 16.11111111111111,
|
| 572 |
+
"grad_norm": 4.259357929229736,
|
| 573 |
+
"learning_rate": 5.150205761316873e-05,
|
| 574 |
+
"loss": 1.6464,
|
| 575 |
+
"step": 2900
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 16.38888888888889,
|
| 579 |
+
"grad_norm": 3.2829339504241943,
|
| 580 |
+
"learning_rate": 5.047325102880659e-05,
|
| 581 |
+
"loss": 1.5819,
|
| 582 |
+
"step": 2950
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 16.666666666666668,
|
| 586 |
+
"grad_norm": 5.9115986824035645,
|
| 587 |
+
"learning_rate": 4.9444444444444446e-05,
|
| 588 |
+
"loss": 1.5292,
|
| 589 |
+
"step": 3000
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 16.944444444444443,
|
| 593 |
+
"grad_norm": 5.184513092041016,
|
| 594 |
+
"learning_rate": 4.841563786008231e-05,
|
| 595 |
+
"loss": 1.5396,
|
| 596 |
+
"step": 3050
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 17.0,
|
| 600 |
+
"eval_accuracy": 0.5612395474667978,
|
| 601 |
+
"eval_f1": 0.38352153766273633,
|
| 602 |
+
"eval_loss": 1.7599705457687378,
|
| 603 |
+
"eval_runtime": 21.2693,
|
| 604 |
+
"eval_samples_per_second": 95.584,
|
| 605 |
+
"eval_steps_per_second": 1.505,
|
| 606 |
+
"step": 3060
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 17.22222222222222,
|
| 610 |
+
"grad_norm": 7.147132396697998,
|
| 611 |
+
"learning_rate": 4.7386831275720164e-05,
|
| 612 |
+
"loss": 1.3781,
|
| 613 |
+
"step": 3100
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 17.5,
|
| 617 |
+
"grad_norm": 4.889599800109863,
|
| 618 |
+
"learning_rate": 4.635802469135803e-05,
|
| 619 |
+
"loss": 1.3611,
|
| 620 |
+
"step": 3150
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 17.77777777777778,
|
| 624 |
+
"grad_norm": 4.039378643035889,
|
| 625 |
+
"learning_rate": 4.532921810699588e-05,
|
| 626 |
+
"loss": 1.3245,
|
| 627 |
+
"step": 3200
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 18.0,
|
| 631 |
+
"eval_accuracy": 0.6040334481062469,
|
| 632 |
+
"eval_f1": 0.44146839497484525,
|
| 633 |
+
"eval_loss": 1.6344797611236572,
|
| 634 |
+
"eval_runtime": 21.2985,
|
| 635 |
+
"eval_samples_per_second": 95.453,
|
| 636 |
+
"eval_steps_per_second": 1.502,
|
| 637 |
+
"step": 3240
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 18.055555555555557,
|
| 641 |
+
"grad_norm": 7.149341583251953,
|
| 642 |
+
"learning_rate": 4.4300411522633745e-05,
|
| 643 |
+
"loss": 1.3637,
|
| 644 |
+
"step": 3250
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 18.333333333333332,
|
| 648 |
+
"grad_norm": 8.064897537231445,
|
| 649 |
+
"learning_rate": 4.327160493827161e-05,
|
| 650 |
+
"loss": 1.2293,
|
| 651 |
+
"step": 3300
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 18.61111111111111,
|
| 655 |
+
"grad_norm": 7.903775691986084,
|
| 656 |
+
"learning_rate": 4.224279835390947e-05,
|
| 657 |
+
"loss": 1.2622,
|
| 658 |
+
"step": 3350
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 18.88888888888889,
|
| 662 |
+
"grad_norm": 4.111396312713623,
|
| 663 |
+
"learning_rate": 4.1213991769547326e-05,
|
| 664 |
+
"loss": 1.1825,
|
| 665 |
+
"step": 3400
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 19.0,
|
| 669 |
+
"eval_accuracy": 0.6246925725528775,
|
| 670 |
+
"eval_f1": 0.4763520182391385,
|
| 671 |
+
"eval_loss": 1.5767754316329956,
|
| 672 |
+
"eval_runtime": 21.0448,
|
| 673 |
+
"eval_samples_per_second": 96.603,
|
| 674 |
+
"eval_steps_per_second": 1.521,
|
| 675 |
+
"step": 3420
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 19.166666666666668,
|
| 679 |
+
"grad_norm": 6.990283966064453,
|
| 680 |
+
"learning_rate": 4.018518518518519e-05,
|
| 681 |
+
"loss": 1.1418,
|
| 682 |
+
"step": 3450
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 19.444444444444443,
|
| 686 |
+
"grad_norm": 5.295820236206055,
|
| 687 |
+
"learning_rate": 3.9156378600823044e-05,
|
| 688 |
+
"loss": 1.1469,
|
| 689 |
+
"step": 3500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 19.72222222222222,
|
| 693 |
+
"grad_norm": 4.929720401763916,
|
| 694 |
+
"learning_rate": 3.812757201646091e-05,
|
| 695 |
+
"loss": 1.0467,
|
| 696 |
+
"step": 3550
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 20.0,
|
| 700 |
+
"grad_norm": 9.373787879943848,
|
| 701 |
+
"learning_rate": 3.709876543209877e-05,
|
| 702 |
+
"loss": 1.0525,
|
| 703 |
+
"step": 3600
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 20.0,
|
| 707 |
+
"eval_accuracy": 0.6374815543531727,
|
| 708 |
+
"eval_f1": 0.49386444314380973,
|
| 709 |
+
"eval_loss": 1.5208549499511719,
|
| 710 |
+
"eval_runtime": 21.2234,
|
| 711 |
+
"eval_samples_per_second": 95.791,
|
| 712 |
+
"eval_steps_per_second": 1.508,
|
| 713 |
+
"step": 3600
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 20.27777777777778,
|
| 717 |
+
"grad_norm": 7.043643474578857,
|
| 718 |
+
"learning_rate": 3.606995884773663e-05,
|
| 719 |
+
"loss": 0.9412,
|
| 720 |
+
"step": 3650
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 20.555555555555557,
|
| 724 |
+
"grad_norm": 4.172306537628174,
|
| 725 |
+
"learning_rate": 3.504115226337449e-05,
|
| 726 |
+
"loss": 0.9267,
|
| 727 |
+
"step": 3700
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 20.833333333333332,
|
| 731 |
+
"grad_norm": 7.322742462158203,
|
| 732 |
+
"learning_rate": 3.401234567901235e-05,
|
| 733 |
+
"loss": 0.9496,
|
| 734 |
+
"step": 3750
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 21.0,
|
| 738 |
+
"eval_accuracy": 0.6586325627151992,
|
| 739 |
+
"eval_f1": 0.526846190918778,
|
| 740 |
+
"eval_loss": 1.4922317266464233,
|
| 741 |
+
"eval_runtime": 21.2176,
|
| 742 |
+
"eval_samples_per_second": 95.816,
|
| 743 |
+
"eval_steps_per_second": 1.508,
|
| 744 |
+
"step": 3780
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 21.11111111111111,
|
| 748 |
+
"grad_norm": 3.450895071029663,
|
| 749 |
+
"learning_rate": 3.2983539094650206e-05,
|
| 750 |
+
"loss": 0.8906,
|
| 751 |
+
"step": 3800
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 21.38888888888889,
|
| 755 |
+
"grad_norm": 4.777499675750732,
|
| 756 |
+
"learning_rate": 3.195473251028807e-05,
|
| 757 |
+
"loss": 0.895,
|
| 758 |
+
"step": 3850
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 21.666666666666668,
|
| 762 |
+
"grad_norm": 6.931682109832764,
|
| 763 |
+
"learning_rate": 3.0925925925925924e-05,
|
| 764 |
+
"loss": 0.8423,
|
| 765 |
+
"step": 3900
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 21.944444444444443,
|
| 769 |
+
"grad_norm": 4.988988876342773,
|
| 770 |
+
"learning_rate": 2.989711934156379e-05,
|
| 771 |
+
"loss": 0.8512,
|
| 772 |
+
"step": 3950
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 22.0,
|
| 776 |
+
"eval_accuracy": 0.6714215445154943,
|
| 777 |
+
"eval_f1": 0.5555183831392889,
|
| 778 |
+
"eval_loss": 1.4699825048446655,
|
| 779 |
+
"eval_runtime": 20.1301,
|
| 780 |
+
"eval_samples_per_second": 100.993,
|
| 781 |
+
"eval_steps_per_second": 1.59,
|
| 782 |
+
"step": 3960
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 22.22222222222222,
|
| 786 |
+
"grad_norm": 5.021641731262207,
|
| 787 |
+
"learning_rate": 2.8888888888888888e-05,
|
| 788 |
+
"loss": 0.8724,
|
| 789 |
+
"step": 4000
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 22.5,
|
| 793 |
+
"grad_norm": 12.957953453063965,
|
| 794 |
+
"learning_rate": 2.786008230452675e-05,
|
| 795 |
+
"loss": 0.7648,
|
| 796 |
+
"step": 4050
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 22.77777777777778,
|
| 800 |
+
"grad_norm": 11.249091148376465,
|
| 801 |
+
"learning_rate": 2.683127572016461e-05,
|
| 802 |
+
"loss": 0.7718,
|
| 803 |
+
"step": 4100
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 23.0,
|
| 807 |
+
"eval_accuracy": 0.6709296606000984,
|
| 808 |
+
"eval_f1": 0.5596452801153066,
|
| 809 |
+
"eval_loss": 1.5005793571472168,
|
| 810 |
+
"eval_runtime": 20.1506,
|
| 811 |
+
"eval_samples_per_second": 100.89,
|
| 812 |
+
"eval_steps_per_second": 1.588,
|
| 813 |
+
"step": 4140
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 23.055555555555557,
|
| 817 |
+
"grad_norm": 6.9278788566589355,
|
| 818 |
+
"learning_rate": 2.580246913580247e-05,
|
| 819 |
+
"loss": 0.7561,
|
| 820 |
+
"step": 4150
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 23.333333333333332,
|
| 824 |
+
"grad_norm": 6.076067924499512,
|
| 825 |
+
"learning_rate": 2.4773662551440328e-05,
|
| 826 |
+
"loss": 0.6688,
|
| 827 |
+
"step": 4200
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 23.61111111111111,
|
| 831 |
+
"grad_norm": 9.00234317779541,
|
| 832 |
+
"learning_rate": 2.374485596707819e-05,
|
| 833 |
+
"loss": 0.6977,
|
| 834 |
+
"step": 4250
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 23.88888888888889,
|
| 838 |
+
"grad_norm": 6.011639595031738,
|
| 839 |
+
"learning_rate": 2.271604938271605e-05,
|
| 840 |
+
"loss": 0.7335,
|
| 841 |
+
"step": 4300
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 24.0,
|
| 845 |
+
"eval_accuracy": 0.6783079193310378,
|
| 846 |
+
"eval_f1": 0.5742235106898269,
|
| 847 |
+
"eval_loss": 1.4700865745544434,
|
| 848 |
+
"eval_runtime": 20.1667,
|
| 849 |
+
"eval_samples_per_second": 100.81,
|
| 850 |
+
"eval_steps_per_second": 1.587,
|
| 851 |
+
"step": 4320
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 24.166666666666668,
|
| 855 |
+
"grad_norm": 6.378417491912842,
|
| 856 |
+
"learning_rate": 2.168724279835391e-05,
|
| 857 |
+
"loss": 0.7581,
|
| 858 |
+
"step": 4350
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 24.444444444444443,
|
| 862 |
+
"grad_norm": 5.407553672790527,
|
| 863 |
+
"learning_rate": 2.0658436213991768e-05,
|
| 864 |
+
"loss": 0.645,
|
| 865 |
+
"step": 4400
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 24.72222222222222,
|
| 869 |
+
"grad_norm": 4.054404258728027,
|
| 870 |
+
"learning_rate": 1.962962962962963e-05,
|
| 871 |
+
"loss": 0.644,
|
| 872 |
+
"step": 4450
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 25.0,
|
| 876 |
+
"grad_norm": 6.489627838134766,
|
| 877 |
+
"learning_rate": 1.860082304526749e-05,
|
| 878 |
+
"loss": 0.6514,
|
| 879 |
+
"step": 4500
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 25.0,
|
| 883 |
+
"eval_accuracy": 0.6787998032464339,
|
| 884 |
+
"eval_f1": 0.5776489069621243,
|
| 885 |
+
"eval_loss": 1.5083099603652954,
|
| 886 |
+
"eval_runtime": 20.1289,
|
| 887 |
+
"eval_samples_per_second": 100.999,
|
| 888 |
+
"eval_steps_per_second": 1.59,
|
| 889 |
+
"step": 4500
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 25.27777777777778,
|
| 893 |
+
"grad_norm": 4.55994176864624,
|
| 894 |
+
"learning_rate": 1.757201646090535e-05,
|
| 895 |
+
"loss": 0.6414,
|
| 896 |
+
"step": 4550
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 25.555555555555557,
|
| 900 |
+
"grad_norm": 3.7992448806762695,
|
| 901 |
+
"learning_rate": 1.654320987654321e-05,
|
| 902 |
+
"loss": 0.572,
|
| 903 |
+
"step": 4600
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 25.833333333333332,
|
| 907 |
+
"grad_norm": 28.1027774810791,
|
| 908 |
+
"learning_rate": 1.551440329218107e-05,
|
| 909 |
+
"loss": 0.6161,
|
| 910 |
+
"step": 4650
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 26.0,
|
| 914 |
+
"eval_accuracy": 0.6802754549926218,
|
| 915 |
+
"eval_f1": 0.5875122879521407,
|
| 916 |
+
"eval_loss": 1.532254695892334,
|
| 917 |
+
"eval_runtime": 20.1215,
|
| 918 |
+
"eval_samples_per_second": 101.036,
|
| 919 |
+
"eval_steps_per_second": 1.59,
|
| 920 |
+
"step": 4680
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 26.11111111111111,
|
| 924 |
+
"grad_norm": 6.135079383850098,
|
| 925 |
+
"learning_rate": 1.448559670781893e-05,
|
| 926 |
+
"loss": 0.6043,
|
| 927 |
+
"step": 4700
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 26.38888888888889,
|
| 931 |
+
"grad_norm": 4.171236515045166,
|
| 932 |
+
"learning_rate": 1.3477366255144033e-05,
|
| 933 |
+
"loss": 0.6035,
|
| 934 |
+
"step": 4750
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 26.666666666666668,
|
| 938 |
+
"grad_norm": 5.243923664093018,
|
| 939 |
+
"learning_rate": 1.2448559670781894e-05,
|
| 940 |
+
"loss": 0.5319,
|
| 941 |
+
"step": 4800
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 26.944444444444443,
|
| 945 |
+
"grad_norm": 7.352174282073975,
|
| 946 |
+
"learning_rate": 1.1419753086419753e-05,
|
| 947 |
+
"loss": 0.5576,
|
| 948 |
+
"step": 4850
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 27.0,
|
| 952 |
+
"eval_accuracy": 0.6822429906542056,
|
| 953 |
+
"eval_f1": 0.5893952239065103,
|
| 954 |
+
"eval_loss": 1.5246381759643555,
|
| 955 |
+
"eval_runtime": 20.121,
|
| 956 |
+
"eval_samples_per_second": 101.039,
|
| 957 |
+
"eval_steps_per_second": 1.59,
|
| 958 |
+
"step": 4860
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 27.22222222222222,
|
| 962 |
+
"grad_norm": 6.190335750579834,
|
| 963 |
+
"learning_rate": 1.0390946502057614e-05,
|
| 964 |
+
"loss": 0.5641,
|
| 965 |
+
"step": 4900
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 27.5,
|
| 969 |
+
"grad_norm": 4.43634557723999,
|
| 970 |
+
"learning_rate": 9.362139917695475e-06,
|
| 971 |
+
"loss": 0.5445,
|
| 972 |
+
"step": 4950
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 27.77777777777778,
|
| 976 |
+
"grad_norm": 3.3575663566589355,
|
| 977 |
+
"learning_rate": 8.333333333333334e-06,
|
| 978 |
+
"loss": 0.5092,
|
| 979 |
+
"step": 5000
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 28.0,
|
| 983 |
+
"eval_accuracy": 0.691096901131333,
|
| 984 |
+
"eval_f1": 0.6042086142382966,
|
| 985 |
+
"eval_loss": 1.5102460384368896,
|
| 986 |
+
"eval_runtime": 20.1291,
|
| 987 |
+
"eval_samples_per_second": 100.998,
|
| 988 |
+
"eval_steps_per_second": 1.59,
|
| 989 |
+
"step": 5040
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 28.055555555555557,
|
| 993 |
+
"grad_norm": 11.934253692626953,
|
| 994 |
+
"learning_rate": 7.304526748971194e-06,
|
| 995 |
+
"loss": 0.5371,
|
| 996 |
+
"step": 5050
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 28.333333333333332,
|
| 1000 |
+
"grad_norm": 5.053288459777832,
|
| 1001 |
+
"learning_rate": 6.275720164609055e-06,
|
| 1002 |
+
"loss": 0.498,
|
| 1003 |
+
"step": 5100
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 28.61111111111111,
|
| 1007 |
+
"grad_norm": 4.996842861175537,
|
| 1008 |
+
"learning_rate": 5.246913580246914e-06,
|
| 1009 |
+
"loss": 0.5505,
|
| 1010 |
+
"step": 5150
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 28.88888888888889,
|
| 1014 |
+
"grad_norm": 13.71835994720459,
|
| 1015 |
+
"learning_rate": 4.218106995884774e-06,
|
| 1016 |
+
"loss": 0.5109,
|
| 1017 |
+
"step": 5200
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 29.0,
|
| 1021 |
+
"eval_accuracy": 0.6886374815543532,
|
| 1022 |
+
"eval_f1": 0.600059116410086,
|
| 1023 |
+
"eval_loss": 1.5202569961547852,
|
| 1024 |
+
"eval_runtime": 20.2693,
|
| 1025 |
+
"eval_samples_per_second": 100.299,
|
| 1026 |
+
"eval_steps_per_second": 1.579,
|
| 1027 |
+
"step": 5220
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 29.166666666666668,
|
| 1031 |
+
"grad_norm": 3.211416721343994,
|
| 1032 |
+
"learning_rate": 3.1893004115226337e-06,
|
| 1033 |
+
"loss": 0.5093,
|
| 1034 |
+
"step": 5250
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 29.444444444444443,
|
| 1038 |
+
"grad_norm": 4.839755058288574,
|
| 1039 |
+
"learning_rate": 2.1604938271604937e-06,
|
| 1040 |
+
"loss": 0.5158,
|
| 1041 |
+
"step": 5300
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 29.72222222222222,
|
| 1045 |
+
"grad_norm": 5.453362464904785,
|
| 1046 |
+
"learning_rate": 1.131687242798354e-06,
|
| 1047 |
+
"loss": 0.522,
|
| 1048 |
+
"step": 5350
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 30.0,
|
| 1052 |
+
"grad_norm": 2.713297128677368,
|
| 1053 |
+
"learning_rate": 1.0288065843621401e-07,
|
| 1054 |
+
"loss": 0.5301,
|
| 1055 |
+
"step": 5400
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 30.0,
|
| 1059 |
+
"eval_accuracy": 0.690605017215937,
|
| 1060 |
+
"eval_f1": 0.603264981616179,
|
| 1061 |
+
"eval_loss": 1.5189286470413208,
|
| 1062 |
+
"eval_runtime": 20.2156,
|
| 1063 |
+
"eval_samples_per_second": 100.566,
|
| 1064 |
+
"eval_steps_per_second": 1.583,
|
| 1065 |
+
"step": 5400
|
| 1066 |
+
}
|
| 1067 |
+
],
|
| 1068 |
+
"logging_steps": 50,
|
| 1069 |
+
"max_steps": 5400,
|
| 1070 |
+
"num_input_tokens_seen": 0,
|
| 1071 |
+
"num_train_epochs": 30,
|
| 1072 |
+
"save_steps": 100,
|
| 1073 |
+
"stateful_callbacks": {
|
| 1074 |
+
"TrainerControl": {
|
| 1075 |
+
"args": {
|
| 1076 |
+
"should_epoch_stop": false,
|
| 1077 |
+
"should_evaluate": false,
|
| 1078 |
+
"should_log": false,
|
| 1079 |
+
"should_save": true,
|
| 1080 |
+
"should_training_stop": true
|
| 1081 |
+
},
|
| 1082 |
+
"attributes": {}
|
| 1083 |
+
}
|
| 1084 |
+
},
|
| 1085 |
+
"total_flos": 1.0469261495692793e+20,
|
| 1086 |
+
"train_batch_size": 64,
|
| 1087 |
+
"trial_name": null,
|
| 1088 |
+
"trial_params": null
|
| 1089 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22e0378208bd6eff3bb7b51946e6fd442631d52fac140fd650cedab09a1a15c5
|
| 3 |
+
size 5240
|