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 +711 -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": "calendar_query",
|
62 |
+
"1": "calendar_set",
|
63 |
+
"10": "iot_hue_lighton",
|
64 |
+
"11": "lists_remove",
|
65 |
+
"12": "lists_query",
|
66 |
+
"13": "play_podcasts",
|
67 |
+
"14": "play_audiobook",
|
68 |
+
"15": "news_query",
|
69 |
+
"16": "iot_wemo_off",
|
70 |
+
"17": "transport_ticket",
|
71 |
+
"18": "audio_volume_down",
|
72 |
+
"19": "calendar_remove",
|
73 |
+
"2": "qa_factoid",
|
74 |
+
"20": "datetime_query",
|
75 |
+
"21": "general_joke",
|
76 |
+
"22": "email_query",
|
77 |
+
"23": "general_greet",
|
78 |
+
"24": "play_game",
|
79 |
+
"25": "audio_volume_mute",
|
80 |
+
"26": "qa_currency",
|
81 |
+
"27": "audio_volume_up",
|
82 |
+
"28": "transport_query",
|
83 |
+
"29": "iot_hue_lightdim",
|
84 |
+
"3": "transport_traffic",
|
85 |
+
"30": "weather_query",
|
86 |
+
"31": "social_post",
|
87 |
+
"32": "qa_definition",
|
88 |
+
"33": "email_querycontact",
|
89 |
+
"34": "takeaway_query",
|
90 |
+
"35": "general_quirky",
|
91 |
+
"36": "iot_cleaning",
|
92 |
+
"37": "play_music",
|
93 |
+
"38": "iot_hue_lightchange",
|
94 |
+
"39": "recommendation_events",
|
95 |
+
"4": "alarm_remove",
|
96 |
+
"40": "iot_coffee",
|
97 |
+
"41": "audio_volume_other",
|
98 |
+
"42": "email_sendemail",
|
99 |
+
"43": "music_settings",
|
100 |
+
"44": "iot_hue_lightoff",
|
101 |
+
"45": "recommendation_movies",
|
102 |
+
"46": "cooking_query",
|
103 |
+
"47": "takeaway_order",
|
104 |
+
"48": "play_radio",
|
105 |
+
"49": "recommendation_locations",
|
106 |
+
"5": "iot_hue_lightup",
|
107 |
+
"50": "music_query",
|
108 |
+
"51": "lists_createoradd",
|
109 |
+
"52": "social_query",
|
110 |
+
"53": "qa_maths",
|
111 |
+
"54": "email_addcontact",
|
112 |
+
"55": "alarm_set",
|
113 |
+
"56": "iot_wemo_on",
|
114 |
+
"57": "transport_taxi",
|
115 |
+
"58": "music_dislikeness",
|
116 |
+
"59": "music_likeness",
|
117 |
+
"6": "qa_stock",
|
118 |
+
"7": "datetime_convert",
|
119 |
+
"8": "alarm_query",
|
120 |
+
"9": "cooking_recipe"
|
121 |
+
},
|
122 |
+
"initializer_range": 0.02,
|
123 |
+
"intermediate_size": 4096,
|
124 |
+
"label2id": {
|
125 |
+
"alarm_query": "8",
|
126 |
+
"alarm_remove": "4",
|
127 |
+
"alarm_set": "55",
|
128 |
+
"audio_volume_down": "18",
|
129 |
+
"audio_volume_mute": "25",
|
130 |
+
"audio_volume_other": "41",
|
131 |
+
"audio_volume_up": "27",
|
132 |
+
"calendar_query": "0",
|
133 |
+
"calendar_remove": "19",
|
134 |
+
"calendar_set": "1",
|
135 |
+
"cooking_query": "46",
|
136 |
+
"cooking_recipe": "9",
|
137 |
+
"datetime_convert": "7",
|
138 |
+
"datetime_query": "20",
|
139 |
+
"email_addcontact": "54",
|
140 |
+
"email_query": "22",
|
141 |
+
"email_querycontact": "33",
|
142 |
+
"email_sendemail": "42",
|
143 |
+
"general_greet": "23",
|
144 |
+
"general_joke": "21",
|
145 |
+
"general_quirky": "35",
|
146 |
+
"iot_cleaning": "36",
|
147 |
+
"iot_coffee": "40",
|
148 |
+
"iot_hue_lightchange": "38",
|
149 |
+
"iot_hue_lightdim": "29",
|
150 |
+
"iot_hue_lightoff": "44",
|
151 |
+
"iot_hue_lighton": "10",
|
152 |
+
"iot_hue_lightup": "5",
|
153 |
+
"iot_wemo_off": "16",
|
154 |
+
"iot_wemo_on": "56",
|
155 |
+
"lists_createoradd": "51",
|
156 |
+
"lists_query": "12",
|
157 |
+
"lists_remove": "11",
|
158 |
+
"music_dislikeness": "58",
|
159 |
+
"music_likeness": "59",
|
160 |
+
"music_query": "50",
|
161 |
+
"music_settings": "43",
|
162 |
+
"news_query": "15",
|
163 |
+
"play_audiobook": "14",
|
164 |
+
"play_game": "24",
|
165 |
+
"play_music": "37",
|
166 |
+
"play_podcasts": "13",
|
167 |
+
"play_radio": "48",
|
168 |
+
"qa_currency": "26",
|
169 |
+
"qa_definition": "32",
|
170 |
+
"qa_factoid": "2",
|
171 |
+
"qa_maths": "53",
|
172 |
+
"qa_stock": "6",
|
173 |
+
"recommendation_events": "39",
|
174 |
+
"recommendation_locations": "49",
|
175 |
+
"recommendation_movies": "45",
|
176 |
+
"social_post": "31",
|
177 |
+
"social_query": "52",
|
178 |
+
"takeaway_order": "47",
|
179 |
+
"takeaway_query": "34",
|
180 |
+
"transport_query": "28",
|
181 |
+
"transport_taxi": "57",
|
182 |
+
"transport_ticket": "17",
|
183 |
+
"transport_traffic": "3",
|
184 |
+
"weather_query": "30"
|
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 sha256:87f6dc37846d81a024a74d3d2f346710b9b15eb9d23c8f2196651722ca1cb988
|
3 |
+
size 1262918936
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c8db0e0f9673b161ce08b3d17d4b0b57ad14a48a61e26a71682cf7f4709e0775
|
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:a93158c9b13d301ebe95b771161f72707903579f86232f8680efad7dc727a13a
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,711 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.5543531726512543,
|
3 |
+
"best_model_checkpoint": "fr-FR/wav2vec2-xls-r-300m/checkpoint-2610",
|
4 |
+
"epoch": 30.0,
|
5 |
+
"eval_steps": 100,
|
6 |
+
"global_step": 2700,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.5555555555555556,
|
13 |
+
"grad_norm": 0.6311665773391724,
|
14 |
+
"learning_rate": 1.814814814814815e-05,
|
15 |
+
"loss": 4.0831,
|
16 |
+
"step": 50
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 1.0,
|
20 |
+
"eval_accuracy": 0.05017215937038859,
|
21 |
+
"eval_f1": 0.0016194974794585797,
|
22 |
+
"eval_loss": 3.921128511428833,
|
23 |
+
"eval_runtime": 20.2481,
|
24 |
+
"eval_samples_per_second": 100.404,
|
25 |
+
"eval_steps_per_second": 1.58,
|
26 |
+
"step": 90
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"epoch": 1.1111111111111112,
|
30 |
+
"grad_norm": 0.6617798805236816,
|
31 |
+
"learning_rate": 3.6666666666666666e-05,
|
32 |
+
"loss": 3.9927,
|
33 |
+
"step": 100
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 1.6666666666666665,
|
37 |
+
"grad_norm": 0.8150258660316467,
|
38 |
+
"learning_rate": 5.518518518518519e-05,
|
39 |
+
"loss": 3.8256,
|
40 |
+
"step": 150
|
41 |
+
},
|
42 |
+
{
|
43 |
+
"epoch": 2.0,
|
44 |
+
"eval_accuracy": 0.06443679291687161,
|
45 |
+
"eval_f1": 0.00205206930041668,
|
46 |
+
"eval_loss": 3.73803973197937,
|
47 |
+
"eval_runtime": 20.3477,
|
48 |
+
"eval_samples_per_second": 99.913,
|
49 |
+
"eval_steps_per_second": 1.573,
|
50 |
+
"step": 180
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 2.2222222222222223,
|
54 |
+
"grad_norm": 0.8914035558700562,
|
55 |
+
"learning_rate": 7.37037037037037e-05,
|
56 |
+
"loss": 3.7727,
|
57 |
+
"step": 200
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 2.7777777777777777,
|
61 |
+
"grad_norm": 0.93168705701828,
|
62 |
+
"learning_rate": 9.222222222222223e-05,
|
63 |
+
"loss": 3.7519,
|
64 |
+
"step": 250
|
65 |
+
},
|
66 |
+
{
|
67 |
+
"epoch": 3.0,
|
68 |
+
"eval_accuracy": 0.07181505164781112,
|
69 |
+
"eval_f1": 0.005259036785425362,
|
70 |
+
"eval_loss": 3.7302937507629395,
|
71 |
+
"eval_runtime": 20.203,
|
72 |
+
"eval_samples_per_second": 100.629,
|
73 |
+
"eval_steps_per_second": 1.584,
|
74 |
+
"step": 270
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 3.3333333333333335,
|
78 |
+
"grad_norm": 0.8782232999801636,
|
79 |
+
"learning_rate": 9.880658436213992e-05,
|
80 |
+
"loss": 3.7506,
|
81 |
+
"step": 300
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 3.888888888888889,
|
85 |
+
"grad_norm": 1.1496119499206543,
|
86 |
+
"learning_rate": 9.674897119341564e-05,
|
87 |
+
"loss": 3.7476,
|
88 |
+
"step": 350
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 4.0,
|
92 |
+
"eval_accuracy": 0.06443679291687161,
|
93 |
+
"eval_f1": 0.00205206930041668,
|
94 |
+
"eval_loss": 3.709669351577759,
|
95 |
+
"eval_runtime": 20.1865,
|
96 |
+
"eval_samples_per_second": 100.711,
|
97 |
+
"eval_steps_per_second": 1.585,
|
98 |
+
"step": 360
|
99 |
+
},
|
100 |
+
{
|
101 |
+
"epoch": 4.444444444444445,
|
102 |
+
"grad_norm": 0.9949362277984619,
|
103 |
+
"learning_rate": 9.469135802469137e-05,
|
104 |
+
"loss": 3.7454,
|
105 |
+
"step": 400
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 5.0,
|
109 |
+
"grad_norm": 1.1786142587661743,
|
110 |
+
"learning_rate": 9.263374485596709e-05,
|
111 |
+
"loss": 3.7098,
|
112 |
+
"step": 450
|
113 |
+
},
|
114 |
+
{
|
115 |
+
"epoch": 5.0,
|
116 |
+
"eval_accuracy": 0.09591736350221347,
|
117 |
+
"eval_f1": 0.005725705609705114,
|
118 |
+
"eval_loss": 3.690065622329712,
|
119 |
+
"eval_runtime": 20.1867,
|
120 |
+
"eval_samples_per_second": 100.71,
|
121 |
+
"eval_steps_per_second": 1.585,
|
122 |
+
"step": 450
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 5.555555555555555,
|
126 |
+
"grad_norm": 1.3047127723693848,
|
127 |
+
"learning_rate": 9.057613168724281e-05,
|
128 |
+
"loss": 3.6858,
|
129 |
+
"step": 500
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 6.0,
|
133 |
+
"eval_accuracy": 0.09886866699458928,
|
134 |
+
"eval_f1": 0.007264866491847572,
|
135 |
+
"eval_loss": 3.6419930458068848,
|
136 |
+
"eval_runtime": 20.0133,
|
137 |
+
"eval_samples_per_second": 101.582,
|
138 |
+
"eval_steps_per_second": 1.599,
|
139 |
+
"step": 540
|
140 |
+
},
|
141 |
+
{
|
142 |
+
"epoch": 6.111111111111111,
|
143 |
+
"grad_norm": 1.4781264066696167,
|
144 |
+
"learning_rate": 8.851851851851852e-05,
|
145 |
+
"loss": 3.6583,
|
146 |
+
"step": 550
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 6.666666666666667,
|
150 |
+
"grad_norm": 1.342982292175293,
|
151 |
+
"learning_rate": 8.646090534979425e-05,
|
152 |
+
"loss": 3.6096,
|
153 |
+
"step": 600
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 7.0,
|
157 |
+
"eval_accuracy": 0.11264141662567634,
|
158 |
+
"eval_f1": 0.012811569793871857,
|
159 |
+
"eval_loss": 3.566227674484253,
|
160 |
+
"eval_runtime": 20.2051,
|
161 |
+
"eval_samples_per_second": 100.618,
|
162 |
+
"eval_steps_per_second": 1.584,
|
163 |
+
"step": 630
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 7.222222222222222,
|
167 |
+
"grad_norm": 1.2319382429122925,
|
168 |
+
"learning_rate": 8.452674897119341e-05,
|
169 |
+
"loss": 3.6427,
|
170 |
+
"step": 650
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 7.777777777777778,
|
174 |
+
"grad_norm": 1.120038628578186,
|
175 |
+
"learning_rate": 8.246913580246915e-05,
|
176 |
+
"loss": 3.5484,
|
177 |
+
"step": 700
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 8.0,
|
181 |
+
"eval_accuracy": 0.12985735366453516,
|
182 |
+
"eval_f1": 0.01639035591982137,
|
183 |
+
"eval_loss": 3.5015017986297607,
|
184 |
+
"eval_runtime": 20.2258,
|
185 |
+
"eval_samples_per_second": 100.515,
|
186 |
+
"eval_steps_per_second": 1.582,
|
187 |
+
"step": 720
|
188 |
+
},
|
189 |
+
{
|
190 |
+
"epoch": 8.333333333333334,
|
191 |
+
"grad_norm": 1.6429734230041504,
|
192 |
+
"learning_rate": 8.041152263374486e-05,
|
193 |
+
"loss": 3.4963,
|
194 |
+
"step": 750
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"epoch": 8.88888888888889,
|
198 |
+
"grad_norm": 2.2033281326293945,
|
199 |
+
"learning_rate": 7.835390946502059e-05,
|
200 |
+
"loss": 3.4439,
|
201 |
+
"step": 800
|
202 |
+
},
|
203 |
+
{
|
204 |
+
"epoch": 9.0,
|
205 |
+
"eval_accuracy": 0.12395474667978357,
|
206 |
+
"eval_f1": 0.019783749206849655,
|
207 |
+
"eval_loss": 3.391706943511963,
|
208 |
+
"eval_runtime": 20.2246,
|
209 |
+
"eval_samples_per_second": 100.521,
|
210 |
+
"eval_steps_per_second": 1.582,
|
211 |
+
"step": 810
|
212 |
+
},
|
213 |
+
{
|
214 |
+
"epoch": 9.444444444444445,
|
215 |
+
"grad_norm": 2.9443519115448,
|
216 |
+
"learning_rate": 7.62962962962963e-05,
|
217 |
+
"loss": 3.3565,
|
218 |
+
"step": 850
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 10.0,
|
222 |
+
"grad_norm": 2.37030029296875,
|
223 |
+
"learning_rate": 7.423868312757202e-05,
|
224 |
+
"loss": 3.2907,
|
225 |
+
"step": 900
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 10.0,
|
229 |
+
"eval_accuracy": 0.1736350221347762,
|
230 |
+
"eval_f1": 0.03961650753233977,
|
231 |
+
"eval_loss": 3.2812328338623047,
|
232 |
+
"eval_runtime": 20.1829,
|
233 |
+
"eval_samples_per_second": 100.729,
|
234 |
+
"eval_steps_per_second": 1.585,
|
235 |
+
"step": 900
|
236 |
+
},
|
237 |
+
{
|
238 |
+
"epoch": 10.555555555555555,
|
239 |
+
"grad_norm": 3.3137433528900146,
|
240 |
+
"learning_rate": 7.218106995884773e-05,
|
241 |
+
"loss": 3.1707,
|
242 |
+
"step": 950
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 11.0,
|
246 |
+
"eval_accuracy": 0.1751106738809641,
|
247 |
+
"eval_f1": 0.04082360136597588,
|
248 |
+
"eval_loss": 3.1738343238830566,
|
249 |
+
"eval_runtime": 20.1847,
|
250 |
+
"eval_samples_per_second": 100.72,
|
251 |
+
"eval_steps_per_second": 1.585,
|
252 |
+
"step": 990
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 11.11111111111111,
|
256 |
+
"grad_norm": 4.893256664276123,
|
257 |
+
"learning_rate": 7.012345679012346e-05,
|
258 |
+
"loss": 3.093,
|
259 |
+
"step": 1000
|
260 |
+
},
|
261 |
+
{
|
262 |
+
"epoch": 11.666666666666666,
|
263 |
+
"grad_norm": 2.7285029888153076,
|
264 |
+
"learning_rate": 6.806584362139918e-05,
|
265 |
+
"loss": 3.0034,
|
266 |
+
"step": 1050
|
267 |
+
},
|
268 |
+
{
|
269 |
+
"epoch": 12.0,
|
270 |
+
"eval_accuracy": 0.23266109198229218,
|
271 |
+
"eval_f1": 0.07458858219754078,
|
272 |
+
"eval_loss": 2.987241744995117,
|
273 |
+
"eval_runtime": 20.172,
|
274 |
+
"eval_samples_per_second": 100.783,
|
275 |
+
"eval_steps_per_second": 1.586,
|
276 |
+
"step": 1080
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 12.222222222222221,
|
280 |
+
"grad_norm": 2.4686148166656494,
|
281 |
+
"learning_rate": 6.60082304526749e-05,
|
282 |
+
"loss": 2.8933,
|
283 |
+
"step": 1100
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 12.777777777777779,
|
287 |
+
"grad_norm": 4.950862884521484,
|
288 |
+
"learning_rate": 6.395061728395062e-05,
|
289 |
+
"loss": 2.7734,
|
290 |
+
"step": 1150
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 13.0,
|
294 |
+
"eval_accuracy": 0.26512543039842595,
|
295 |
+
"eval_f1": 0.09626588280325074,
|
296 |
+
"eval_loss": 2.8068132400512695,
|
297 |
+
"eval_runtime": 20.1739,
|
298 |
+
"eval_samples_per_second": 100.774,
|
299 |
+
"eval_steps_per_second": 1.586,
|
300 |
+
"step": 1170
|
301 |
+
},
|
302 |
+
{
|
303 |
+
"epoch": 13.333333333333334,
|
304 |
+
"grad_norm": 3.508140802383423,
|
305 |
+
"learning_rate": 6.189300411522633e-05,
|
306 |
+
"loss": 2.6421,
|
307 |
+
"step": 1200
|
308 |
+
},
|
309 |
+
{
|
310 |
+
"epoch": 13.88888888888889,
|
311 |
+
"grad_norm": 2.7219550609588623,
|
312 |
+
"learning_rate": 5.983539094650206e-05,
|
313 |
+
"loss": 2.5912,
|
314 |
+
"step": 1250
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 14.0,
|
318 |
+
"eval_accuracy": 0.2951303492375799,
|
319 |
+
"eval_f1": 0.12066140989612666,
|
320 |
+
"eval_loss": 2.687779188156128,
|
321 |
+
"eval_runtime": 20.1345,
|
322 |
+
"eval_samples_per_second": 100.971,
|
323 |
+
"eval_steps_per_second": 1.589,
|
324 |
+
"step": 1260
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 14.444444444444445,
|
328 |
+
"grad_norm": 3.374047040939331,
|
329 |
+
"learning_rate": 5.7777777777777776e-05,
|
330 |
+
"loss": 2.4899,
|
331 |
+
"step": 1300
|
332 |
+
},
|
333 |
+
{
|
334 |
+
"epoch": 15.0,
|
335 |
+
"grad_norm": 4.324727535247803,
|
336 |
+
"learning_rate": 5.57201646090535e-05,
|
337 |
+
"loss": 2.4839,
|
338 |
+
"step": 1350
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 15.0,
|
342 |
+
"eval_accuracy": 0.32611903590752583,
|
343 |
+
"eval_f1": 0.16030752546112262,
|
344 |
+
"eval_loss": 2.6126816272735596,
|
345 |
+
"eval_runtime": 20.1788,
|
346 |
+
"eval_samples_per_second": 100.749,
|
347 |
+
"eval_steps_per_second": 1.586,
|
348 |
+
"step": 1350
|
349 |
+
},
|
350 |
+
{
|
351 |
+
"epoch": 15.555555555555555,
|
352 |
+
"grad_norm": 3.545776128768921,
|
353 |
+
"learning_rate": 5.366255144032922e-05,
|
354 |
+
"loss": 2.3293,
|
355 |
+
"step": 1400
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"epoch": 16.0,
|
359 |
+
"eval_accuracy": 0.3453025086079685,
|
360 |
+
"eval_f1": 0.17236182167002215,
|
361 |
+
"eval_loss": 2.5199337005615234,
|
362 |
+
"eval_runtime": 20.1733,
|
363 |
+
"eval_samples_per_second": 100.777,
|
364 |
+
"eval_steps_per_second": 1.586,
|
365 |
+
"step": 1440
|
366 |
+
},
|
367 |
+
{
|
368 |
+
"epoch": 16.11111111111111,
|
369 |
+
"grad_norm": 5.616283893585205,
|
370 |
+
"learning_rate": 5.160493827160494e-05,
|
371 |
+
"loss": 2.2367,
|
372 |
+
"step": 1450
|
373 |
+
},
|
374 |
+
{
|
375 |
+
"epoch": 16.666666666666668,
|
376 |
+
"grad_norm": 5.327425479888916,
|
377 |
+
"learning_rate": 4.9547325102880656e-05,
|
378 |
+
"loss": 2.169,
|
379 |
+
"step": 1500
|
380 |
+
},
|
381 |
+
{
|
382 |
+
"epoch": 17.0,
|
383 |
+
"eval_accuracy": 0.36399409739301525,
|
384 |
+
"eval_f1": 0.2078663996139009,
|
385 |
+
"eval_loss": 2.4475929737091064,
|
386 |
+
"eval_runtime": 20.1551,
|
387 |
+
"eval_samples_per_second": 100.868,
|
388 |
+
"eval_steps_per_second": 1.588,
|
389 |
+
"step": 1530
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 17.22222222222222,
|
393 |
+
"grad_norm": 5.141805171966553,
|
394 |
+
"learning_rate": 4.748971193415638e-05,
|
395 |
+
"loss": 2.1224,
|
396 |
+
"step": 1550
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 17.77777777777778,
|
400 |
+
"grad_norm": 3.864804744720459,
|
401 |
+
"learning_rate": 4.54320987654321e-05,
|
402 |
+
"loss": 2.0474,
|
403 |
+
"step": 1600
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 18.0,
|
407 |
+
"eval_accuracy": 0.3954746679783571,
|
408 |
+
"eval_f1": 0.2329830407945391,
|
409 |
+
"eval_loss": 2.3817105293273926,
|
410 |
+
"eval_runtime": 20.1551,
|
411 |
+
"eval_samples_per_second": 100.868,
|
412 |
+
"eval_steps_per_second": 1.588,
|
413 |
+
"step": 1620
|
414 |
+
},
|
415 |
+
{
|
416 |
+
"epoch": 18.333333333333332,
|
417 |
+
"grad_norm": 5.070396423339844,
|
418 |
+
"learning_rate": 4.337448559670782e-05,
|
419 |
+
"loss": 1.986,
|
420 |
+
"step": 1650
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 18.88888888888889,
|
424 |
+
"grad_norm": 3.915616035461426,
|
425 |
+
"learning_rate": 4.1316872427983537e-05,
|
426 |
+
"loss": 1.9385,
|
427 |
+
"step": 1700
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"epoch": 19.0,
|
431 |
+
"eval_accuracy": 0.43433349729463844,
|
432 |
+
"eval_f1": 0.26422232789593036,
|
433 |
+
"eval_loss": 2.2753679752349854,
|
434 |
+
"eval_runtime": 20.1569,
|
435 |
+
"eval_samples_per_second": 100.859,
|
436 |
+
"eval_steps_per_second": 1.588,
|
437 |
+
"step": 1710
|
438 |
+
},
|
439 |
+
{
|
440 |
+
"epoch": 19.444444444444443,
|
441 |
+
"grad_norm": 4.9511399269104,
|
442 |
+
"learning_rate": 3.925925925925926e-05,
|
443 |
+
"loss": 1.833,
|
444 |
+
"step": 1750
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 20.0,
|
448 |
+
"grad_norm": 8.35456657409668,
|
449 |
+
"learning_rate": 3.720164609053498e-05,
|
450 |
+
"loss": 1.8169,
|
451 |
+
"step": 1800
|
452 |
+
},
|
453 |
+
{
|
454 |
+
"epoch": 20.0,
|
455 |
+
"eval_accuracy": 0.45696015740285295,
|
456 |
+
"eval_f1": 0.3069655313049267,
|
457 |
+
"eval_loss": 2.233362913131714,
|
458 |
+
"eval_runtime": 20.1609,
|
459 |
+
"eval_samples_per_second": 100.839,
|
460 |
+
"eval_steps_per_second": 1.587,
|
461 |
+
"step": 1800
|
462 |
+
},
|
463 |
+
{
|
464 |
+
"epoch": 20.555555555555557,
|
465 |
+
"grad_norm": 4.3977885246276855,
|
466 |
+
"learning_rate": 3.51440329218107e-05,
|
467 |
+
"loss": 1.7037,
|
468 |
+
"step": 1850
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"epoch": 21.0,
|
472 |
+
"eval_accuracy": 0.4864731923266109,
|
473 |
+
"eval_f1": 0.32092427700938486,
|
474 |
+
"eval_loss": 2.1784141063690186,
|
475 |
+
"eval_runtime": 20.146,
|
476 |
+
"eval_samples_per_second": 100.913,
|
477 |
+
"eval_steps_per_second": 1.588,
|
478 |
+
"step": 1890
|
479 |
+
},
|
480 |
+
{
|
481 |
+
"epoch": 21.11111111111111,
|
482 |
+
"grad_norm": 4.727329730987549,
|
483 |
+
"learning_rate": 3.308641975308642e-05,
|
484 |
+
"loss": 1.6356,
|
485 |
+
"step": 1900
|
486 |
+
},
|
487 |
+
{
|
488 |
+
"epoch": 21.666666666666668,
|
489 |
+
"grad_norm": 8.652691841125488,
|
490 |
+
"learning_rate": 3.102880658436214e-05,
|
491 |
+
"loss": 1.5998,
|
492 |
+
"step": 1950
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 22.0,
|
496 |
+
"eval_accuracy": 0.5081160846040335,
|
497 |
+
"eval_f1": 0.3454217533127742,
|
498 |
+
"eval_loss": 2.1258203983306885,
|
499 |
+
"eval_runtime": 20.1442,
|
500 |
+
"eval_samples_per_second": 100.922,
|
501 |
+
"eval_steps_per_second": 1.589,
|
502 |
+
"step": 1980
|
503 |
+
},
|
504 |
+
{
|
505 |
+
"epoch": 22.22222222222222,
|
506 |
+
"grad_norm": 10.133247375488281,
|
507 |
+
"learning_rate": 2.897119341563786e-05,
|
508 |
+
"loss": 1.549,
|
509 |
+
"step": 2000
|
510 |
+
},
|
511 |
+
{
|
512 |
+
"epoch": 22.77777777777778,
|
513 |
+
"grad_norm": 4.033301830291748,
|
514 |
+
"learning_rate": 2.6913580246913585e-05,
|
515 |
+
"loss": 1.51,
|
516 |
+
"step": 2050
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 23.0,
|
520 |
+
"eval_accuracy": 0.5081160846040335,
|
521 |
+
"eval_f1": 0.3574234127712544,
|
522 |
+
"eval_loss": 2.0932371616363525,
|
523 |
+
"eval_runtime": 20.1291,
|
524 |
+
"eval_samples_per_second": 100.998,
|
525 |
+
"eval_steps_per_second": 1.59,
|
526 |
+
"step": 2070
|
527 |
+
},
|
528 |
+
{
|
529 |
+
"epoch": 23.333333333333332,
|
530 |
+
"grad_norm": 3.8187780380249023,
|
531 |
+
"learning_rate": 2.4855967078189303e-05,
|
532 |
+
"loss": 1.4493,
|
533 |
+
"step": 2100
|
534 |
+
},
|
535 |
+
{
|
536 |
+
"epoch": 23.88888888888889,
|
537 |
+
"grad_norm": 3.596712589263916,
|
538 |
+
"learning_rate": 2.279835390946502e-05,
|
539 |
+
"loss": 1.4407,
|
540 |
+
"step": 2150
|
541 |
+
},
|
542 |
+
{
|
543 |
+
"epoch": 24.0,
|
544 |
+
"eval_accuracy": 0.5164781111657649,
|
545 |
+
"eval_f1": 0.37192680141832524,
|
546 |
+
"eval_loss": 2.098966121673584,
|
547 |
+
"eval_runtime": 20.1169,
|
548 |
+
"eval_samples_per_second": 101.059,
|
549 |
+
"eval_steps_per_second": 1.591,
|
550 |
+
"step": 2160
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 24.444444444444443,
|
554 |
+
"grad_norm": 4.555119037628174,
|
555 |
+
"learning_rate": 2.074074074074074e-05,
|
556 |
+
"loss": 1.415,
|
557 |
+
"step": 2200
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 25.0,
|
561 |
+
"grad_norm": 4.245804786682129,
|
562 |
+
"learning_rate": 1.868312757201646e-05,
|
563 |
+
"loss": 1.3468,
|
564 |
+
"step": 2250
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 25.0,
|
568 |
+
"eval_accuracy": 0.5366453516969996,
|
569 |
+
"eval_f1": 0.38983496988780386,
|
570 |
+
"eval_loss": 2.0432238578796387,
|
571 |
+
"eval_runtime": 20.0898,
|
572 |
+
"eval_samples_per_second": 101.195,
|
573 |
+
"eval_steps_per_second": 1.593,
|
574 |
+
"step": 2250
|
575 |
+
},
|
576 |
+
{
|
577 |
+
"epoch": 25.555555555555557,
|
578 |
+
"grad_norm": 4.591546058654785,
|
579 |
+
"learning_rate": 1.662551440329218e-05,
|
580 |
+
"loss": 1.2885,
|
581 |
+
"step": 2300
|
582 |
+
},
|
583 |
+
{
|
584 |
+
"epoch": 26.0,
|
585 |
+
"eval_accuracy": 0.5391047712739794,
|
586 |
+
"eval_f1": 0.40537387724487417,
|
587 |
+
"eval_loss": 2.01814866065979,
|
588 |
+
"eval_runtime": 20.1408,
|
589 |
+
"eval_samples_per_second": 100.939,
|
590 |
+
"eval_steps_per_second": 1.589,
|
591 |
+
"step": 2340
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 26.11111111111111,
|
595 |
+
"grad_norm": 3.095262050628662,
|
596 |
+
"learning_rate": 1.4567901234567902e-05,
|
597 |
+
"loss": 1.2853,
|
598 |
+
"step": 2350
|
599 |
+
},
|
600 |
+
{
|
601 |
+
"epoch": 26.666666666666668,
|
602 |
+
"grad_norm": 3.832343339920044,
|
603 |
+
"learning_rate": 1.2510288065843623e-05,
|
604 |
+
"loss": 1.2539,
|
605 |
+
"step": 2400
|
606 |
+
},
|
607 |
+
{
|
608 |
+
"epoch": 27.0,
|
609 |
+
"eval_accuracy": 0.5479586817511067,
|
610 |
+
"eval_f1": 0.41006063615437155,
|
611 |
+
"eval_loss": 2.021588087081909,
|
612 |
+
"eval_runtime": 20.1524,
|
613 |
+
"eval_samples_per_second": 100.881,
|
614 |
+
"eval_steps_per_second": 1.588,
|
615 |
+
"step": 2430
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 27.22222222222222,
|
619 |
+
"grad_norm": 4.070197105407715,
|
620 |
+
"learning_rate": 1.0452674897119343e-05,
|
621 |
+
"loss": 1.2327,
|
622 |
+
"step": 2450
|
623 |
+
},
|
624 |
+
{
|
625 |
+
"epoch": 27.77777777777778,
|
626 |
+
"grad_norm": 3.290456533432007,
|
627 |
+
"learning_rate": 8.395061728395062e-06,
|
628 |
+
"loss": 1.21,
|
629 |
+
"step": 2500
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 28.0,
|
633 |
+
"eval_accuracy": 0.5484505656665027,
|
634 |
+
"eval_f1": 0.41472217066308065,
|
635 |
+
"eval_loss": 2.0198214054107666,
|
636 |
+
"eval_runtime": 20.1563,
|
637 |
+
"eval_samples_per_second": 100.862,
|
638 |
+
"eval_steps_per_second": 1.588,
|
639 |
+
"step": 2520
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 28.333333333333332,
|
643 |
+
"grad_norm": 3.003391981124878,
|
644 |
+
"learning_rate": 6.337448559670782e-06,
|
645 |
+
"loss": 1.2188,
|
646 |
+
"step": 2550
|
647 |
+
},
|
648 |
+
{
|
649 |
+
"epoch": 28.88888888888889,
|
650 |
+
"grad_norm": 5.390565395355225,
|
651 |
+
"learning_rate": 4.2798353909465025e-06,
|
652 |
+
"loss": 1.1995,
|
653 |
+
"step": 2600
|
654 |
+
},
|
655 |
+
{
|
656 |
+
"epoch": 29.0,
|
657 |
+
"eval_accuracy": 0.5543531726512543,
|
658 |
+
"eval_f1": 0.4205878734286696,
|
659 |
+
"eval_loss": 2.0205445289611816,
|
660 |
+
"eval_runtime": 20.1392,
|
661 |
+
"eval_samples_per_second": 100.947,
|
662 |
+
"eval_steps_per_second": 1.589,
|
663 |
+
"step": 2610
|
664 |
+
},
|
665 |
+
{
|
666 |
+
"epoch": 29.444444444444443,
|
667 |
+
"grad_norm": 3.9864342212677,
|
668 |
+
"learning_rate": 2.2222222222222225e-06,
|
669 |
+
"loss": 1.197,
|
670 |
+
"step": 2650
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"epoch": 30.0,
|
674 |
+
"grad_norm": 3.8882803916931152,
|
675 |
+
"learning_rate": 1.646090534979424e-07,
|
676 |
+
"loss": 1.1697,
|
677 |
+
"step": 2700
|
678 |
+
},
|
679 |
+
{
|
680 |
+
"epoch": 30.0,
|
681 |
+
"eval_accuracy": 0.5528775209050664,
|
682 |
+
"eval_f1": 0.4185233692128861,
|
683 |
+
"eval_loss": 2.020263910293579,
|
684 |
+
"eval_runtime": 20.0868,
|
685 |
+
"eval_samples_per_second": 101.211,
|
686 |
+
"eval_steps_per_second": 1.593,
|
687 |
+
"step": 2700
|
688 |
+
}
|
689 |
+
],
|
690 |
+
"logging_steps": 50,
|
691 |
+
"max_steps": 2700,
|
692 |
+
"num_input_tokens_seen": 0,
|
693 |
+
"num_train_epochs": 30,
|
694 |
+
"save_steps": 100,
|
695 |
+
"stateful_callbacks": {
|
696 |
+
"TrainerControl": {
|
697 |
+
"args": {
|
698 |
+
"should_epoch_stop": false,
|
699 |
+
"should_evaluate": false,
|
700 |
+
"should_log": false,
|
701 |
+
"should_save": true,
|
702 |
+
"should_training_stop": true
|
703 |
+
},
|
704 |
+
"attributes": {}
|
705 |
+
}
|
706 |
+
},
|
707 |
+
"total_flos": 1.0469261495692793e+20,
|
708 |
+
"train_batch_size": 64,
|
709 |
+
"trial_name": null,
|
710 |
+
"trial_params": null
|
711 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4dfc098e2f42f7078edb1b5cd5ff5649939c31582cbdd97e3d65511426a6e66f
|
3 |
+
size 5240
|